How can I get my 12ax7's to distort as good as transistors?

[guitician]

Ok, so let me start by saying that I LOVE tubes for giving my electric guitar that "Just Right" tone and feel when I play it.

But when I am working with 12ax7 circuits I've noticed that they don't always cooperate in the distortion dept. By that I mean, when the signal into a series of 12ax7's is brought to the point of clipping, the sound is "hit or miss". I'm working on an old amp... http://music-electronics-forum.com/t32108/ and it had a really nice distortion with it's stages arranged in way that was different from the MKI boogies, but the reverb's output was really low. So, in the process of getting the reverb to work right; by making it a near exact clone of the MKI(voltages and circuit paths the same), the distortion tone now blows.

The four series stages are all on the same power node and have the usual 100k/1.5k plate/cathode arrangement. The cathodes are bypassed with 47uf 50v caps, not the 22uf that the boogie used, which would give it more low freq. gain. but before the tone was great with the 47uf's so I don't think that's a problem. I haven't done tube swaps because the tone was fine before with the tubes that are in it. The one thing I did that may be an issue is add 270k ohm in series with the grid on the second stage like the boogie...and move the tone control circuit from ahead of the last stage, to after the second.

I'm half tempted just to play it clean, which sounds fantastic, and use a distortion box instead......

[Chuck H]

To be fair...

If you jimmy with any SS device that is intended to create good distortion, with the intent of improving some different aspect, you will likely detriment the quality of the distortion.?. Too ethereal???

Whatever makes good tube distortion doesn't necessarily make for good tube reverb. By that I mean to say that an amp that can do both must be hybridized for that specific purpose. To take an amp that is not idealized for either and tweak for one OR the other will always mean that the "other" is going to be lacking. Evidenced by the fact that there are very many all tube guitar amps, of different topologies, that create very desirable distortion. Everything from uber gain to classic rock to rockabilly!!! And, of course, reverb is a tube amp staple. The two can share space, but only if designed that way. Which can get quite complicated with an all tube design (just ask ME, I've solved for this very problem). My two cents is that if you dig the sound of certain D-boxes, use them. There is NO shame in that. I would even venture to say that MOST of the great distortion tones prior to 1998 (and even many thereafter) included the use of distortion pedals. Most often a combination of tube clipping and the D-box contribution.

OK... I'm all over the map here. But let me qualify... Tubes are no more limited than transistors in their ability to make good clipped tones when combined with reverb effects. Quite the contrary. But the circuit designs are somewhat prohibitive. Most manufacturers opt for a hybrid circuit. USING THE TUBES FOR THE CLIPPING!!! And run the reverb circuit with SS devices. As a home brewer/diy your limited only by the amount of knowledge and effort your willing or able to invest.

If you can examine the circuit you just changed and isolate what it was about that circuit that made the clipped tone good, you may be able to preserve it with alternative corrections to the lacking reverb circuit. Unless the amp, as a platform, doesn't afford such gratuity.

I hope this wasn't to ambiguous.

[Steve Conner]

You turned it into a Boogie clone and the distortion tone was bad? I think that is self-explanatory

[guitician]

Well. I think the Mitchell DeLuxe was already a mesa clone of sorts. I don't have an original schematic and the circuit was modded I think by switching the positions of the stages, which are all connected through shielded cables to controls on a separate chromed metal panel to resemble an old fender. But the chassis is missing one side and has this panel screwed to it; kind of a joke if you ask me. I think my problem could be that the tone controls were moved "up front" and are now picking up stray hum fields because of the un-sheiled wires I used on them.

And the tube vs. SS distortion issue was getting at my relative ease at getting really good sounding distortion with SS vs. 12ax7's

[daz]

My theory, and i am adamant about this, is that tube distortion is NOT what makes tubes so great. It's tube dynamics....feel, etc. SS devices typically are far superior at shaping distortion and you can do a lot more and far easier. Look at the array of SS pedals and how widely the tones vary. To get that kind of shaping out of 12AX7 circuits you would have to have all sorts of garbage going on and you still wouldn't get there. And as far as you CAN go there, theres no answer to your question because it would be a huge number of things you'd have to do.

I've mentioned this quote often, but it rings so true to me i think it's important. Aspen Pittman former owner of GT once said, and i'm paraphrasing, "it's not that tubes sound better than SS, it's that they feel better. IMO, and i've proven this to myself thru years of experimentation, that tube dynamic goodness (NOT talking tube distortion) exists primarily in the power section not the pre. So in short, your best bet is simply to use a super good and super clean tube amp and pedals. A lot of the best players on the planets do just that. Personally the one problem i have with that approach is i relay heavily on the style of playing where you use the guitar to control your tone and clean to mean, and pedals don't do that the way even cascaded 12AX7's can assuming the design is great. In fact, thats what i spent years trying to do with my homebrew and succeeded to a great extent. (no small thanks to some members here) But in any case, you aren't going to get the kind of tonal variation from 12ax7's you're looking for. They are a one trick pony with some variation but nothing like transistors are capable of.

[guitician]

Yeah Daz, I hear ya... the dynamics of how tubes open up, and compress, while being played provides a feeling of having more control over what is happening.
I've made my SS boxes do just that, and they really make me believe I'm playing tubes.
This amps circuit at the moment has no trick, other then clean/dynamic. But I've heard some MKI boogie and Mitchell Pro-100 stuff that really does sound like I would want this amp to sound like, so I guess I'm gonna keep plugging at it until I get some "trick" distortion.

[defaced]

If you liked the distortion before fixing the reverb, then change it back and find a different solution to the reverb problem. Making the reverb output louder could be as simple as inserting another stage, be it tube or MOSFET. The LND150 would be my go to device for this if you determine you need voltage gain.

[guitician]

fixing the reverb issue was a matter of adding the 2.2Mohm/20pf deal at the last gain stage before the phase inverter. Then balancing the ratio of reverb to dry signal. I really dont need to mess with something that is a "standard" setup.

It's just that the tube stages need some sort of logical sequencing that envelopes the signal and makes the distortion "sing". Right now it's "clipping" all right, but there's a "trick" missing. I looked at how Soldano bias's cold/hot and uses two VF's too, so I think maybe biasing for the "overdriven" signal may be the way to go.

[Chuck H]

OK. But not all amps that clip well use the Soldono circuit. And not all amps would benefit from it. Just as the big resistor bypassed by a cap isn't used in all amps with reverb. And not all amps will benefit from it. If you added it then it's not a standard circuit for that amp. So the consideration that you don't need to mess with it because it's a standard circuit is only going to snafu your efforts. I said earlier that you should identify what it was that made the distortion better and restore that. Then find an alternative solution for the reverb problem. This sentiment was echoed by Mike.

EDIT: I don't know how close the Mitchell amp actually was to the Mesa schematic you posted. There may be some significant differences that make referencing it moot for the purposes of this discussion. For example, the Mesa DOES use a 2.2M/10pf circuit for the reverb. The Mitchell didn't. When it was added to the Mitchell it didn't work. Why? I dunno. I don't have a Mitchell schematic for comparison.

[guitician]

I think the Mitchell did use the 2.2M/20pf because there is a trace for a parallel small component with a resistor at the grid of the last stage and it seemed to make sense when I saw how the MKI did it.
I really don't remember how i had it wired before to tell the truth... and the fact that one 12ax7 is being used on either side of the 2.2M/20pf to get a send/return for reverb doesn't seem like it would mess with the distortion much anyway. I think it was when I positioned the tone control to the first stage is where it went bad in the distortion dept. I really need to wire the tone controls with shielded wire.

[Steve Conner]

First of all, if you've added a 20pF cap to the signal path to boost highs, there's a possibility that you might have made the amp unstable. The crappy tone could be from oscillations.

Now, when cascading tubes to get a high-gain, saturated distortion tone, I always found that it turned to crap unless I did some of the following things:

Undersized coupling caps to roll off bass, and speed recovery from blocking distortion.
Undersized cathode bypass caps in the later stages or even none at all.
Attenuators between stages to dump a bit of gain. (Maybe make one of them variable for a gain control.)

Technically, dynamics are a form of distortion. An ideal amp should amplify all signals the same. If it treats loud signals different to quiet ones, that's distortion.

I gave up on ultra-high gain and channel switching. I now do what Daz does, set the amp for a medium amount of gain and use the guitar volume and pickup switches to control it. For some solos, I use a clean boost pedal to abuse the amp some more.

[defaced]

fixing the reverb issue was a matter of adding the 2.2Mohm/20pf deal at the last gain stage before the phase inverter. Then balancing the ratio of reverb to dry signal. I really dont need to mess with something that is a "standard" setup.

It's just that the tube stages need some sort of logical sequencing that envelopes the signal and makes the distortion "sing". Right now it's "clipping" all right, but there's a "trick" missing. I looked at how Soldano bias's cold/hot and uses two VF's too, so I think maybe biasing for the "overdriven" signal may be the way to go.

Unfortunately the details of how you remedied the reverb were not in your first post, so I didn't have much to work from.

"Singing" distortion doesn't really describe much. Singing to me certainly won't be singing to you. Steve has outlined the basic guidelines of building a high gain amp, but there are alot of ways to skin that cat. Some don't have a CF at all (Engl), some have lots of interstate filtering (Diezel), some have no cold stage (I've seen this most on modded amps, it gives more of a saturated tone whereas a cold stage gives more of a raw/open tone (at least in the SLO circuit)). High gain amps are all I build (my current projects are a tube and LND150 based Avenger (SLO OD channel)), so I'm adequately familiar with the territory, but they are picky.

I think it was when I positioned the tone control to the first stage is where it went bad in the distortion dept. I really need to wire the tone controls with shielded wire.

That costs you alot of signal voltage putting the tone stack there and steals your ability to shape the sound of the clipped signal. There's a reason the Mark series amps have a GEQ after all the clipping, it's because a TS that early in a cascaded gain stage amp doesn't do much. Most high gain amps have one of two TS setups: TS early and a GEQ after the preamp, or a single TS after the preamp. I'd move your TS back in the preamp. Also, the earlier in the signal path, the more likely you are to have to shield. I typically don't use shielded wire after the second gain stage. By then the signal is large enough to not really be bothered by noise, but your amp may be more picky (layout counts big time here).

If after you re-locate the TS and still want to play, the biggest thing that sticks out at me in the schematic in that other post are the 22u cathode bypass caps. I'd move them to be closer to 1u as needed to remove some low end and tighten up the sound of the amp.

[Chuck H]

Not to be a stick in the mud, but... How can we give any advice when we really don't know much about the circuit!?! The photo isn't giving as much information as a schematic would. In fact, it only served to make me question the layout. But I will second Mike on one issue that is probably at the root of your problem. Moving the tone stack to the front end sacrifices some overall gain. It also strips your ability to control the EQ post clipping. Did you notice that the EQ no longer does much to the frequency response when the amp is clipping??? To add, there is one other thing that moving the tone stack to the front end does. It drastically changes the frequency setup for later clipping stages. Where a small-ish coupling cap is flat with some LF roll off, a tone stack is scooped in the mids and generally has a lot more bottom end unless you turn the bass down to aesthetically unpleasing levels. So there is a tendency to drive too much LF through the clipping stages. Also, looking at your layout I think it's possible that moving stages around by simply re-routing wires may detriment the circuits stability. As Steve pointed out.

[guitician]

Yeah, stability has been an issue, I had to deal with oscillations when the gain is turned up. When I reduced the sensitivity of the PI a lot of the oscillations ceased. That's why I threw up that photo, so you could see what I'm dealing with.

The amp has four 12ax7's to work with, and with the PI and reverb using two there isn't much "High Gain" really from two 12ax7's with a TS in between. Basically, the three stages played from input #1, do great. So, input #2 should just be like adding a clean boost pedal as Steve suggests. Because of the fact that it has a master volume, and makes the pre-amp "Drivable", there's a "Gain" of sorts. I like the fact that the four stages are all the same (100k/1.5k||47uf - plate/cathode||cap) arrangement because of it's simplicity, and I can use whatever one I see fit, but Mitchell put the inputs on the right side of the amp, which is the same side as the output! I put the reverbs stages closest to the power supply(left side), and then the 1st&2nd stage tube followed by 3rd&4th, and finally PI & power amp. The TS components are on the PC board after the 2nd tube from the left, so to put them later in the circuit I have to swap tubes 2&3 in the wiring.

By "Singing" I meant sustaining a harmonically enriched note that will self-perpetuate. Yeah, the tone controls really don't do much for the driven tone, but have to be set carefully when a driven tone is dialed in. And then, getting back to a nice rhythm tone is a time consuming affair. Which is why this amp is a "one trick pony", that does either one or the other in a given session, and I'm not expecting it to do both. I'm just surprised that moving identical stages did so much change to the sound. Maybe the diifferences in the tubes are to blame...I'll swap them and see.

[guitician]

Most high gain amps have one of two TS setups: TS early and a GEQ after the preamp, or a single TS after the preamp....

If after you re-locate the TS and still want to play, the biggest thing that sticks out at me in the schematic in that other post are the 22u cathode bypass caps. I'd move them to be closer to 1u as needed to remove some low end and tighten up the sound of the amp.

I'm thinking that I can try to tone shape the #2 inputs(boost stage) by using a different bypass cap in series with a resistor, and putting a parallel cap across the plate resistor. Like a fixed tone control so I'm driving the right frequencies.