Loose low end

[daz]

You probably missed my post, but at one point (i believe in my last PSU thread) i said that i was wrong in thinking the PI didn't have a filter. It was just that i had the wires that connected the nodes under the board and the way i had the dropping resistors oriented it looked like the PI had no filter, but it did. And now it's got an even bigger 80uf cap. (22uf before)

As for needing counseling, i didn't need you to tell me that.

[Steve Conner]

Sorry, I edited my last post to include stuff about the bass response of the room you're doing your tweaking in. Could that possibly be it? Or have you tried the amp in other rooms and it sounds crappy everywhere?

It just got me thinking, because if I only ever tweaked amps in my living room, I'd be totally unhappy with the loose, flabby bass. But if I take them through to the bedroom or kitchen the bass tightens up! My living room has brick walls and two dimensions the same, so it's a seriously bad bass situation. The other rooms have sheetrock walls on timber studs, that vibrate a little and absorb the bass resonances.

But sometimes you just need to accept that your work is never going to sound perfect no matter how hard you tweak. You have to draw a line in the sand and say you're just going to play it and enjoy it anyway. It's never too late, there could even be an Ampaholics Anonymous in your town!

Unless of course you enjoy modding more than playing, in which case ignore me

[guitician]

Well, I would replace those 80uf's on the V1&V2 stages with 10-16uf's. Your rolloff of the highs is bringing up the bass, relative to the whole frequency response.

[daz]

Well, I would replace those 80uf's on the V1&V2 stages with 10-16uf's. Your rolloff of the highs is bringing up the bass, relative to the whole frequency response.

I just changed them to 80uf from 22uf that have been in it since day one. The 80's actually helped the issue.

Steve....actually, yes, i DO enjoy tweaking it more than playing ! My left hand went south on me several years ago and i no longer gig and don't play for enjoyment that often. I'm doing this mainly as a nervous habit i guess, because i have no need for an amp at this late point in my life. But having spent so many years chasing the perfect tone i guess whether i need it or not no longer matters...now it's just winning the lifelong battle ! But yes, even tho it's more a case of enjoying the chase, i DO play it here and there and want it to be *right* to my ear for both reasons. Oh, and i frucking nuts too....that helps.

[guitician]

It seems like your raising all the caps values at the same time. Did you try to raise the poweramp caps while lowering the Pre's caps? I saw in your PSU thread how you equate more filtering with more highs, but that may not be true across the board.

[daz]

I've already tried 22's again, and in fact i think i like a 22uf on V1A. But in any case, it's not the caps, or at least not sizes.What i keep trying to say but no one believes me is that this isn't a component value issue/tweak....it's a design issue.

for the record i also went back to fixed bias.....for about 2 minutes ! yep....it's still doesn't sound as good as cathode bias generally, and as low end goes no better.

[octal]

It sounds like the motorboating and the bass issue are probably related.

Some thoughts:

Anything which rolls off the bass response INSIDE the feedback loop will reduce the stability margin of the amp. (low quality OPT, too small coupling caps to the output tubes, etc.) I think in an earlier thread you said you still had the problem with the NFB disconnected, right? If so, we can rule this out.

Any chance of a high resistance ground anywhere?
Any chance you made a mistake about the value of a part?

I could be talking smack here, but I wonder if it would be worthwhile to isolate the plate supplies of V1A and V1B with another power supply node.

You could also try running the power supply nodes parallel rather than series... I.e. start from your screen supply- run a resistor and a filter cap to the PI. Start from the screen supply, run a resistor and a filter cap to the V2 stage. etc. You will have to juggle the value of your dropping resistors of course.

Good luck!

[daz]

In fact i started a thread about parallel nodes because i was interested in that. But i realized thats not likely to do it and too much trouble.

[Chuck H]

Daz, your my kind of crazy if it helps. I went through my worst compulsive streak when I was 22. If I had to do it today I think that level of focus and obsession would kill me. So kudos. I understand perfectly that where your head is creates both a great deal of gratification and some personal sacrifice all at the same time. It's a worthy quest. I emerged from mine as my own personal hero and that's a pretty good feeling.

About the circuit... IMHE the big plate resistors on 12ax7's are good for gain and make for a smoother, fuzzier low end. You aren't getting much OD from V1a to V1b but I'm just pointing it out. Also IMHE, largish cathode resistors create a cold clipping situation. This also makes for a smooth fuzzy bottom end. And the standard Marshall type 820 ohm resistor on the cathode that feeds the cathode follower puts a hot biased preamp stage late in the circuit. This makes for a very non linear saturation that can also take the definition out of the low end. Lastly (is that even a word?) having a bit of upper mid punch seems to add snap to the lower notes without taking away from the cleanish tones.

Below are some mods (shown in red) I think you should try. They are minor changes and I think keep within the spirit of your design.

The classic "Marshall" tone (if thats even what you're after) to my ears actually doesn't have a tight bass. But rather a snappy upper midrange that helps define the low notes. When listening the distinction can be subtle. But it's significant because if you miss whats really happening and try to design in certain properties while sacrificing others (it's always a give and take) you can lose what you were after in the first place.

It shouldn't take much to just try these mods.

Chuck

[daz]

LOL!!! I beat you to a lot of that chuck ! I changed the 220k plate to 100k a few days ago. Thats what i had when i first built it and when i went to 220k i'd occasionally try a 100k again but it would always sound better with the 220k. well, after the PSU mods i've been doing a lot of things are changing as far as what works and what doesn't, so i tried the 100k again and this time i think it took....i'm liking it. Then theres the 220k you proposed in place of the 100k grid leak....changed that to 680k today ! ok, so not to the values you suggested, but i did TRY a 220k and a 330k, (didn't have any 470k's left) and the 680 seemed best as fat as changing the tone in a good way w/o adding too much gain. the second grid leak....i tried it but i went back to the 100k because anything higher gave me the kind of gain that doesn't clean up well. But i didn't go lower because usually i find on these grid leaks if i go under 100k or thereabouts the tone just kinda goes limp. But i'll try a 68k and see. The 1.5k on V2a cathode, i have tried different values there before, but not for a long time. ad the only reason i have a 1uf there is i didn't have any more .68uf's and i needed one for the amp's sibling that my friend has so i took it out and put it in his because in that position 1uf vs .68 didn't really matter due to the small resistor. maybe with the 1.5k it will matter a touch more. i try the 1.5k and a .68 if i can find another one laying around.

I'm not really after classic marshall tone....i'm after what i would say is MY idealized version of marshall tone. By the way, you mentioned going thur a similar compulsive streak at 22.....well i'm just a bit late at 55 ! : And by the way i go at this stuff like a rabid schnauzer and with my lack of articulation you probably thought i was 22, eh? But the level of focus and obsession doesn't kill me because at my age anything that keeps my focus away from death is a good thing.

[daz]

Chuck, you magnificent bastage !!! One of those tweaks or the combo of the ones i did really helped ! The overall tone is better, and the lows still have the issue but are better. Enough so that i can be happy as is even if i don't fix it 100%. The only changes i used are the 1.5k on V2's cathode and the 68k grid leak, but they really helped. That 680k i mentioned i had put as a grid leak where you suggested a 220k....well, i had forgotten but it was NOT there. I had tried it and liked it but forgot that later in the day i have put the 100k back because i had too much gain like that. So i tried the 220k but i'm back to the 100k again. I seem to gravitate towards that because for one thing it always brings more gain than i want and other tweaks to relieve gain never get it down enough. then the tone also starts changing too much. I also didn't do the 1.5k you suggested for V2B's cathode. I tried going lower a few times already and got too much gain again. but i will try that today anyways because with your other changes it may work. I'll let you know. thanks chuck, your ideas definatly helped, and every little bit gets me closer to nirvana.

EDIT: been playing this thing all day and i gotta say, those two simple changes, the 68k grid leak and the 1.5k in place of the 820R on V2 cathode were the magic bullet. It's just so close to perfection now. This amp get some of the best tone i have ever heard and felt, and those two things basically balanced things just right. i now have killer tone and the ability to dial in 2 very different but great traditional sort of tone. one very marshall all the way, and one that sounds like a high gain fender. Not of course close as the marshall tone but awesome. i know theres no way i'll ever stop effing with this thing, but i'll tell you, at this point i absolutely do not need to change a thing and i could be happy with this amp for life. thanks chuck....That was all it took to hit that magic balance where everything is just right. Bass still isn't as tight as some amps, but then i don't want moder tight, so this is actuall very good as is.

EDIT2: it keeps getting better. I spent the entire nite playing it in a mix and it's just freakin' great. The tone is not only as close to perfection as i've had, but in a mix it cuts thru so right. Thats the real test. I'm truly stoked. I put a 68k in the other grid leak position by the way. This amp is really killin'.

[Chuck H]

Your not using a whole lotta gain it seems. Not the uber distorto two bricks cracking together thing anyhow.

I'm glad your loving it. But if I may... The mods I suggested are more about putting the bias point of the tubes into a more linear region. When you bias cold (as in V1b) or hot (as V2A was with the 820R) you can lose definition. Especially on the bottom end. So it's not really about gain structure. Gain can be managed in other ways and IMHO the bias point of the tubes is a bad place to do it. But it's a great place to introduce intentional ditortion characteristics.

I know there have been other changes since your last schem but I used it as a platform anyhow.

The drawing shows the V1b cathode as 1.5k (changes in red as always) and a 100k pot where your split plate load was. The 1.5k cathode R will put this stage into a more linear operation and should further improve bottom end clarity. You can then use the 100k pot to dial in the amount of gain you want from that stage. The pot is temporary and should not be left in there. Once you find the gain you want, just remove the pot (without turning it!!!) and measure from each tab to center to find new fixed resistor values for the split plate load.

Just a suggestion... If you do get more bottom end clarity you may find it possible to drive a little more gain through the circuit for those times when you just want to go to 11.

You can experiment with characteristics that different preamp tube bias resistors offer, but it's usually better to use a load R (the 68k resistors your using) or a split plate load (the two 56k resistors on V1b plate). So now you can control both character AND gain instead of being stuck in a give and take.

And see the note on V2A cathode circuit.

Cheers

Chuck

[daz]

Oops, yes, it's a .1uf on the V2A cathode, not 1uf. Sorry. Anyways, dammit Chuck....must you give me more things to try AFTER i was totally happy with it?! Seriously tho, i figured i have no desire to mess wth it any more it sounds so good, so today for the first time in weeks (except days where i couldn't) i didn't touch it....just played. I'm sorta afraid to do anything else to it for fear i will find something i think sounds better only to find down the road that things just don't sound as good as i thought and then can't get it back to where it was. But i'll look at your ideas more and think about it. The lows are great now. i don't know that i would want them any tighter because i like it so much as is. Everything is just exceptional. As to gain amount....when i say "high gain", to me high gain is about what a peavey classic 30 has on it's disto channel, or a jcm 800. I hate uber high gain. I just like the amount needed for classic rock. I consider that high gain. To me boogie gain and that sort of thing is horrible. My most important goal, and one this amp has now met beautifully is to be able to go from the amp's max gain to a very clean sound when the guitar is on 5 with a single coil and picking softly. Thats the type of clean i like, not CLEAN clean. Dynamic tone that goes from clean to pretty darn grindy with just pick attack. had it before, but now it's even better.

I will put up an updated schematic after i post this reply, so it'll be up within a couple minutes if anyone happens to be reading this right after i hit submit. I will probably try your suggestions new Chuck, because you know me....can't put the d@mn iron down. Thanks again Chuck. I appreciate your help a LOT. You made my day big time with the last suggestions. I'll of course let u know if i try these new ones and what happens.

Heres the new schematic. refresh your browser if you viewed my last one or thats what may show up if you don't.

http://i24.photobucket.com/albums/c3.../myEL34new.jpg

You did it again chuck. Awesome. i didn't end up with your exact values, but you were right in lowering the value of that 4.7k cathode. I tried the 1.5k as you said and a pot as plate load, but no setting was really as good as the amp was before. then i went back to a split load on V1B and a 2.7k on the cathode. The tone remains like it did before except for one thing....it now has this amazing light spongy feel and tone that is even better. Thing is, i have literally tried almost every V1A and B cathode resistor and cap values imaginable in the past, but they never did this. i know why tho. it's because till i recently made great strides in getting the PSU right, the places like this in the amp that are ripe for tweaking were now available to be tweaked. but when the PSU wasn't right it seems only the values i had before worked out well. Chuck, thanks once again. I fear you will suggest other things now !!! I'm so happy with the amp at the moment that for the first time it has more than met the hopes i had for it when i first began this journey. If you lived nearby i'd buy you a few beers. Thanks so much. I seriously have not owned a better amp in my almost 40 years of playing. By the way, if you have any suggestions for a fender style 6L6 amp with a preamp like a blues junior as far as gain amount, but with much better overall tone,(think vibroverb with some nice preamp OD) lemmie know. I may now embark on a new journey. Hey, where you all going? Come back here.....i have questions.

[Chuck H]

Daz,

I'm not going to suggest anything else... Unless you ask of course.

I didn't know that you went with a straight 100k plate R for V1b instead of the 56k/56k split load. Thats why I thought you had reduced gain by changing to the 68k load R. But with the 100k plate R it all makes sense now. Personally I love split loads for gain management. Any time a circuit already has a load (ie: volume pot, tone stack, etc.) if I need to reduce gain I use a split plate to keep the ground loads down to one per gain stage. This helps to minimize phase shifts and sounds better (to my ears) than a large series resistor.

My last post was really just a tutorial on gain management vs bias. I thought it might get you closer to your goals and give you info for any later designing. It's great to intentionally skew the bias of a preamp stage IF it gets the effect you want. But I've seen (and even done it myself) a few tweaks that are intended to reduce gain that actually cool the bias. Since this can make a tube distort in a different, and more non linear way than a more nuetral bias point a bias point should be chosen to get the response you want, not to increase or decrease gain.

So if you like the sound of the amp with the 4.7k bias R on V1b then use it. If you like the feel of the 2.7k great. If you only changed that resistor you should have noticed a slight gain increase. So all the info above and in the last post gets to this... Choose the resistor you think sounds best and adjust the gain with your load R or a split plate load.

For some reason I just realized... I'm out of beer. Someone must have mentioned it.

Cheers

Chuck

P.S. Right now you have the bigger power tube cathode bias bypass cap on the side with the most drive. You could maintain the imbalance but possibly tighten up bass response by swapping sides on the power tube cathode bypass caps... And I said I wouldn't suggest anything

[daz]

Actually chuck, i just put the cap values on the power tube cathodes on the schematic randomly not thinking it mattered. So i'll have to look at the amp and determine which side the 47uf is. As for these tweaks, i think i have to step away for a while because i just tried playing in a CD mix and i couldn't get the nice well mixed punchy tone i got last nite. So i changed it back to how it was before the last mod you showed me and even then it wasn't happening 1 My ears are probably burned out. so i need to step away for a while i think, then try again both ways and see which is best. not sure, but i think the latest tweaks, while giving that great soft feel, it just didn't have the cutting edge to make it thru the mix. Of course the crazy thoughts now start coming and i start thinking i hope i didn't upset the fine balance somehow with the way i soldered them in or some off the wall thing like that.

EDIT: spaced while telling you about the cap on V2a...i meant to say it IS a 1uf, not a .1uf. Brain cluttered with component values, sorry. I also need to keep a log i guess because i am losing track of the values i had in it at various times. i stuck a 2.2k or 2.7k in the V2A cathode position at one point because your 1.5k worked so well i figured why not dump even more gain, and thats what may have been in it last nite when i was really digging it. So i stuck that back in at the end of this nite and will have to wait to see how it does. In any case, i'm at a point now where the tone has gotten fantastic, but i'm needing to determine where to draw the line between killer tone by itself and what works in a mix. We both know those two tones can be very different. then theres the never ending placebo effect and the ear burnout effect, and the weather, and ....etc etc etc....you know what i mean ! But i'm always real close now, and thats the important thing.

EDIT2: today i tried the smaller resistor on V1B again, trying a 1.5k, and 2.7k, and a 3.3k. But the 4.7k sounds best. It really does for some reason. I used split loads on one or 2 of the plates depending on what the cathode R was. For example, with the 1.5k i have split loads on both plates because 1.5k required i take the gain down a good bit. I suppose i could have gone with say a 22k and 82k as the split load, but i wonder if it's better to split both plates with 50/50% value resistors than to split just one with a very small one and large one. In any case, the 4.7k for some reason gives me a fuller and less hard/bright sound and even gets harmonically richer with high OD. From what you said this shouldn't be the better resistor value there, but it really does just plain sound better, subjectivity aside. I did go back to 100k/82k for the PI balance because i A/B'd it and it now sounds better like that. I also have a 2.7k on V2A because i figured if 1.5k made such a nice difference, why not reduce gain there even more. been back and fourth between the 1.5 and 2.7 several times to be sure, and it seems the 2.7k will be staying.