What's happening in this plate circuit?

[Wilder Amplification]

This pretty much sums up the history of guitar amplification, I'd say.

Because ya can't calculate tone. Lots of people in the hi-fi and pro audio community have become "spec monkeys" and think "If this gear sounds this way then any gear with these specs will also sound this way" and due to many exceptions and variables they don't always hold true. Different companies use different methods for arriving at their final specs...such as power amp companies who rate their max power with a burst frequency and arrive at some max power spec that's more than what the amp can possibly draw.

Groover is doing a good thing though by getting into the engineering side of things and putting himself through the math involved, load lines, etc etc...which will teach him how to apply the equations so that when he does do a "trial and error" assessment of what sounds best, he'll be able to perform the math required to reverse engineer what he just did in hopes of gaining an understanding of not only what goes on in different circuits, but also learn why certain circuit topologies sound the way they do as well as figure out what frequency ranges his ears are best tuned to.

[Groover]

....as well as figure out what frequency ranges his ears are best tuned to.

Well if we can measure the frequency response of my tinnitus we'll know the answer to that one at least!

[octal]

The other thing to consider here is how the changing load the output tubes present / reflect back to the driver/PI transformer affects the clipping.

You've got a transformer driving what is essentially an open circuit (the grids of the output tubes) until they hit grid conduction at 0V and start drawing current... when I was playing around with this type of circuit, I found that the size of the grid stopper resistors on the output tubes (I added some...did not directly connect the X-mer to the grids) made a significant difference in the clipping behavior.

You could also try adding a resistive load on the output side of the transformer...would probably clean it up a bit.

Nathan

[Groover]

Thanks again for all the input.

FWIW I am will not be depending on this amp for generating much overdrive/distortion, just to get a [!subjectivity alert!] good clean tone. It would be nice had a little give & fatness to it, but I will be using it with pedals for crunch and lead tones in the intended application.

But this seems so involved I am starting to think that I might be better off converting the amp to the Princeton non-rev aa964 circuit, except for keeping the PA cathode biased.

Of course if the collective wisdom thinks there is something to be gained from sticking with the PIT I'll keep tweaking. OT saturation seems to be desirable to some in some amps. Perhaps it would be in this PIT? Of course I have no idea whether the 12ax7 stage is capable of driving it that hard.

[hasserl]

Well Groover, I'll tell you what. I have two of these little MMB amps. They are not stock, I've modified the Vol/Tone circuit like the tweed Harvard, and bumped up the lone coupling cap to .022 IIRC. And I have to say these little amps are some of the best sounding amps I've owned or played, and I have a prettty good collection of vintage Fenders (Princeton Reverb, Bassman and Super Reverb, a 5E5A and a modified 5E3X2) and about a dozen other amps, plus those I've modified for and those I've built for customers. I'm not a newbie around good tube amps, and they all have good points to them, but I rate these MMB's as excellent sounding amps. I get PLENTY of distortion out of these amps, no problem, but they also sound very nice set clean, or at that "edge of distortion" setting where you can roll the guitar volume to go between a nice spanky clean to a good dirty grind, then hit the front end with a boost or an OD for really nice sustainy feedbacky distortion rock & roll. Great little amps. I don't know why you're doing all these things to try to get something else out of it. Simply rewire the Vol/Tone, play with that 2nd stage plate load a little bit, and you've got an ass kickin little amp that's great for practice volumes. Oh, and ditch that POS speaker, I've found the Weber Alnico Signature 12 is a good inexpensive speaker that mates up perfectly with these amps. In other words, I wouldn't be converting that amp to some bastardized Princeton contraption, the basic MMB is a far better amp IMO. And I do think that trans PI is key to the performance, I wouldn't ditch it either. In fact, if you do let me know how much you want for it, I'll buy it off you and use it in that 4x6v6 1x15" combo I want to build.

[Groover]

My dear Hasserl;

I did initially rewire my MMB with a Tweed Harvard front end like yours, and it did sound very good. I never played it much above 3 or 4, because at that point it was already getting pretty loud, although still fairly clean. Also, I had it mounted in a Vibro Champ cab and I didn't have an appropriate 8" 8ohm speaker to put in there. So it was never a fully constituted self contained amp, I always needed to run it through some available cabinet.

My reason for grafting it with the Vibro Champ front end was to get a full tone stack and vibrato. The situations I want to use the amp for are church gigs and theater pit band gigs where volume must be kept low and space is at a premium. The VC was never quite enough power. So I took everything from the power supply through the PI from the MMB chassis and put it in the VC chassis. I have installed a 10" speaker in the VC cab, and it is serving the intended purpose fairly well.

However when I combined it with the VC front end I had much less available gain/volume. Whereas with the Tweed front end I didn't want to hear it above 4 or so, with the VC front end I need to turn it up to around 7, even with the (added) 25k midrange pot up pretty far, to get a usable volume. I know that the TMB stack is more loss-y than the tweed tone control, but I didn't realize how much so until I did the conversion. Also the preamp tubes are getting lower voltage supply than they did with the original VC PT. So I am just trying to get back a little more gain, and I want to make sure the preamp and vibrato are optimized for the supply voltages they are getting now.

Is that so wrong?

But I do appreciate your sharing your experience with the MMB, and I am glad you have positive things to say about the amp. If I do end up going with the Princeton circuit, I can always re-construct the the MMB/Harvard configuration because we both know that sounds very good.

And I don't think I'll be selling it.

[imaradiostar]

Now that you've explained yourself I have a solution.

Replace the 12ax7 on the front end (both halves) with a 6GH8A or 6U8 or some other kind of triode/pentode with relatively low anode resistance on both sides. Most of those dual tubes are excellent quality and relatively inexpensive. The filament current is only a touch higher than a 12ax7 so you shouldn't have to worry about that.

You'd only have to add a single resistor and cap for the screen grid of the pentode- use a .1uf cap and probably sub a few resistor values till you find something that sounds nice. Depending on your own preference you could use the triode or the pentode as the first stage. Both halves of the aforementioned tubes have low enough plate resistance to work well that way and may even sound better with lower plate resistor and cathode resistor values.

I have been planning (but haven't had time) to build a little combo that is as I've described- triode out front, tone stack (probably trainwreckish values), pentode into phase split TF, then an output section with two or four 6aq5's- not yet decided. I bought a bunch of 6005w's on ebay a while back and I think they sound great. For a similar tone with really cheap tubes you could try those little Russian 9-pin 6v6 equivalent tubes- 6P1P I think.

It might even but interesting to use a dual pot at the grids of the output tubes as a master volume to coax a little more "transformer phase inverter" tone out of it.

jamie

[Groover]

I was thinking about just adding a 7 pin triode for an extra gain stage - or maybe a cathode follower after the first stage to drive the tone stack, but I like your idea because it doesn't require adding a socket.

I looked at the 6GH8A data sheet briefly and the triode section has an amplification factor of only 47, compared to the 12AX7 amplification factor of 100, so unless I am missing something I would be getting way less gain out of the first stage. Perhaps there is a triode/pentode with a triode section closer in performance to the 12AX7.

How would I go about finding what all the tube types are that combine a triode and a pentode in one 9pin envelope? Is there a table or something that lists tube #s according to type? As near as I can tell there is not a direct systematic relationship between the type of valves within a tube and the tube designation #. Maybe there is and I haven't recognized the pattern.

[Steve Conner]

Try searching on the European part numbers. Unlike the American ones they're actually meaningful. All triode/power pentode combos begin in ECL, and triode/signal pentode combos are ECF.

The ECL82 and ECL86 are the most popular ones, they both have American part numbers too.

[Groover]

Please clarify what is meant by power pentode vs signal pentode.

[Steve Conner]

It's the difference between an EL84 and an EF86. Power pentodes are optimized for generating lots of audio power. Signal ones are optimized for high gain, low noise and low microphonics, but can't deliver more than a few milliwatts.

[imaradiostar]

I was thinking about just adding a 7 pin triode for an extra gain stage - or maybe a cathode follower after the first stage to drive the tone stack, but I like your idea because it doesn't require adding a socket.

I looked at the 6GH8A data sheet briefly and the triode section has an amplification factor of only 47, compared to the 12AX7 amplification factor of 100, so unless I am missing something I would be getting way less gain out of the first stage. Perhaps there is a triode/pentode with a triode section closer in performance to the 12AX7.

How would I go about finding what all the tube types are that combine a triode and a pentode in one 9pin envelope? Is there a table or something that lists tube #s according to type? As near as I can tell there is not a direct systematic relationship between the type of valves within a tube and the tube designation #. Maybe there is and I haven't recognized the pattern.

There is a lot more to gain than just mu- it's a case of doing work into a load. Used to drive the transformer in parallel with a 15k resistor a 12ax7 has an actual voltage gain of around 17- it just doesn't have the balls to drive the 15k resistive load in parallel with the transformer (50k AC load at a wild guess). The 6GH8A triode has an actual in-circuit gain of about 32- quite a bit more- enough to make up for a lot of the loss in the tone stack.

This of course doesn't account for the added gain of the pentode which could be between 75 and 125 depending on a number of things. I can say with a decent amount of confidence that even though I haven't built it, it'll work pretty well.

jamie

[Groover]

To clarify, you are suggesting the triode as first stage and pentode as second stage, or vice~versa?

Will the tremolo injection into the 2nd stage cathode circuit still work well?

I notice that the 6GH8 is referred to as a triode/sharp-cutoff pentode. It is a signal pentode? I assume this means it will have a more abrupt cutoff point, and harder clipping on that side of the waveform. Any implications for implementation other than that?

[imaradiostar]

I don't fully understand why they call it "sharp cutoff" but I believe it has to do with how the grid curves all bunch up on the datasheets. Long story short- the gain is relatively constant over a variety of currents. A semi-remote or remote cutoff pentode will have gain that varies over its operating range. This leads to vari-mu or similar compressor designs. I think the Fairchild limiters of Beetles fame used some sort of remote cutoff pentode as a gain control element.

Clipping isn't determined by the grid curves or "sharp cutoff" designation- clipping is determined by how you bias the tube. The 6GH8a has a much lower anode resistance than most classic "audio" preamp pentodes like the EF86. You could safely go with between 15k and 47k plate resistance with the 6GH8a pentode and put it closer to its "sweet spot." It'll draw a lot more current than an EF86 but it'll do a lot more work too. If you were using an EF86 you'd be hard pressed to drive a tone stack without gain loss. The 6GH8 or 6U8 will laugh at most tone stacks because their anode resistance is so much lower- while still having gain well above that of a 12ax7.

You might wanna look at Merlin's Valve Wizard site and possibly even buy his book. It's a lot to learn but it'll equip you with the skills to make educated choices and modify your amplifiers to sound the way you like them to, or at least know why some don't sound good!

Here's the important thing to remember for tube audio- it doesn't have to say "audio" on the datasheet for it to be a fabulous audio tube. The 6L6 was largely a derivative of a RF transmitter tube. The russian version we like so much these days is (if I've read correctly) a servo tube for aircraft use.

I think Steve was trying to describe that tubes can be used in various ways with no trouble. Maybe he was saying the opposite, I don't know!

The ECL86 or 6GW8 is a good example. They are more or less identical- euro and US designations. My RCA tube manual says it's a sharp cutoff pentode and a high gain triode. European datasheets call it a triode-output pentode. You can use it either way depending on your mood and how you connect it and where it is in the circuit.

Many millions of interesting but somewhat obscure TV tubes were made and now sit collecting dust. I'm an advocate of finding ones that are musically useful like the 6GH8a or 6GN8. There are literally hundreds so grab an old tube manual and start reading- it'll blow your mind!

You can arrange the pentode and triode however you like. I only used the triode as the second stage because it made gain calculations easier (no screen grid to account for) and I knew the triode had high transconductance, meaning it could easily drive the low load presented by the transformer and the 15k resistor. While it may not have the gain of the 12ax7, it can create more current change for a given amount of voltage change at the input.

In fact- if you have the filament current available, a 6GN8 may be a good one for you to try as well. The triode section is more like a 12ax7 so it could be on the input and drive the tone stack and the pentode is capable of 5 or so watts of output- enough drive to really abuse the output tubes if that's what you want.

You have a push-pull output stage. You should take advantage of that and inject the trem at the center tap of the phase splitter transformer. Look at the vibro-champ schematic. Connect the center tap of your phase splitter transformer's secondary to the non-grounded end of the 25k "trem" pot from the vibrochamp schematic. This assumes both output tubes are pretty closely matched. Some claim this is the best kind of trem in Fender-land. See if you agree. You can always add an opto-isolator and make it look more like a lot of other "ground shunt" fender trem circuits out there. I personally would keep the trem away from the tube driving the PI transformer- that sucker needs to push pretty hard so it's set up for a lower impedance, making it harder to wiggle the cathode with the trem circuit. Not saying it won't work- just saying you might have better results and an easier time varying the output section's bias.

Hope I didn't overload you too much! Just solder some things up and listen, see if you like them or not! Just keep one hand in your pocket if you're probing.

jamie

[Groover]

Jaimie;

Thank you so much. I have Merlin's book and have read through most of it once, but it will take some study to really digest.

Thanks for the suggestion about connecting the PIT CT to the tremolo. I will try that before I do anything else. - Would I still use both 1/2s of the 12ax7 - the oscillator side and the cathode follower or whatever it is side?

And I may try the pentode in the first stage. I have a lot of options to try now when I get home tomorrow.

Thanks again for all your help.