What's happening in this plate circuit?

[Steve Conner]

If you're using one of the ECF tubes, you could try the pentode in the first stage and the triode as the driver. The 6GH8a is similar to the ECF82, and this is how it was used in Dynaco amps: the pentode as the input stage and the triode as a phase inverter.

But if it's an ECL type, I think the triode should be used as the first stage, and the pentode used to drive the interstage transformer, because this is how they're designed to be used: triode for preamplification, pentode for power amplification.

Use them the other way round at your own risk, because power pentodes aren't specified for low noise, and signal ones aren't specified for high output power. You found that the 12AX7 is marginal as an interstage driver, the triode section of an ECL tube or the pentode section of an ECF will probably be no better, because they'll be like the 12AX7 and EF86 respectively.


Maybe you could get interesting results by triode connecting the pentode. I'd definitely recommend that if you were using an ECL type tube with the pentode driving the interstage. The triode connection lowers the plate impedance drastically, which will help to damp the interstage transformer.

The Fender 300/400PS are the only other amps I know of that used an interstage. The 300PS drives it with a triode connected 6V6, the 400PS uses a triode connected 6L6.


There's also the 12DW7/ECC832/ECC823. Half of a 12AX7 and half of a 12AU7 in the same bottle. JJ ECC832 / 12DW7

Might be an interesting drop-in upgrade for the MM Bass: I imagine the 12AU7 section could drive the interstage transformer a lot harder. The ECC823 is a reverse version.

[octal]

The transformer PI circuit I played around with used a 12AU7 to drive the transformer and I think I had a single half of a 12AX7A as the front end...that was enough gain to get into overdrive (using a simple treble cut tone stack.) So Steve's suggestion to try a 12DW7 might be a good one.

The turns ratio of your PI transformer has a lot to do with the total circuit gain as well. I used one from tubesandmore.com (Antique Electronic Supply.) The P/N is P-T156 , it has a 1:3 turns ratio, so figuring the center tap will effectively halve the turns ratio that should be a 1.5X voltage gain. It's a tiny little X-mer and costs about $15. I wonder what the stock turns ratio on the MMB PI transformer is?

Nathan

[Groover]

Looks like the "12AU7" side is wired to pins 123 so as in the MMB you would need to switch the wiring of the V1 tube socket around to get the respecitve triodes where you wanted them in the circuit.

So it's not quite a "drop-in" upgrade, in case anyone wants to try it.

[Groover]

Can I use a +4dbu tone to test the transformer for voltage ratio, or do I need a stronger signal? I have a test tone generator that will give me +4 , 440Hz or 1KHz.
I have read that someone on this forum uses 120 mains current, but that would be my 2nd choice for safety reasons.

[imaradiostar]

With 7.55 volts across the MMB interstage primary the voltage across the entire secondary is 9.9 volts- 4.95 between the center tap and each grid connection. This means for every 1.5 volts into the primary, each grid moves 1 volt. It's about a 1:1.3 ratio overall.

As Steve mentioned above you can get a 12dw7 with the sections reversed. Look for the JJ ecc823 and 832. I don't recall which is which.

I took a few more minutes and did some calculating. Try a 12at7 in the circuit the way you have it. The at7 will have about the same gain as the ax7 in the first stage (give or take a dB or so) but substantially more gain than an au7 or ax7 in the second position. It'll increase the gain of the second stage by about 15- up from 17 to a little over 30.

Some say the 12at7 doesn't distort as musically as other tubes. In this circuit I doubt it'll matter- you'll mainly be hitting the output tubes harder. The 12at7 should make up for the added loss in the tone stack.

jamie

[Groover]

With 7.55 volts across the MMB interstage primary the voltage across the entire secondary is 9.9 volts- 4.95 between the center tap and each grid connection. This means for every 1.5 volts into the primary, each grid moves 1 volt. It's about a 1:1.3 ratio overall.

Thanks for that.

As Steve mentioned above you can get a 12dw7 with the sections reversed. Look for the JJ ecc823 and 832. I don't recall which is which.

I couldn't find any ECC823 tubes online, but I was thinking if I reverse the 9 pin socket 180 degrees and rewire it, It will put the "AX7" triode of the EC832 in the V1a position and the "AU7" triode in the V1b position, making it easy to swap back to a symmetrical dual triode.

I took a few more minutes and did some calculating. Try a 12at7 in the circuit the way you have it. The at7 will have about the same gain as the ax7 in the first stage (give or take a dB or so) but substantially more gain than an au7 or ax7 in the second position. It'll increase the gain of the second stage by about 15- up from 17 to a little over 30.

Some say the 12at7 doesn't distort as musically as other tubes. In this circuit I doubt it'll matter- you'll mainly be hitting the output tubes harder. The 12at7 should make up for the added loss in the tone stack.

It sounds as if you are saying I can sub in the 12AT7 without changing any plate or cathode resistors. Is that correct? I'll try it as soon as I have time.

Also please note my question in a previous post about tremolo injection - do I leave cathode follower section of VC tremolo intact when I wire the PIT secondary to the intensity pot? I guess I can try it and see without blowing anything up.

[imaradiostar]

Keep the cathode follower- less to modify and it lowers the output impedance of the trem.

jamie

[Groover]

I changed the 2nd resistor in the filter string from 10k to 1k to up the preamp voltage. Now V1a & V1b are getting +276 V up from +256 v. I moved the vibrato B+ to come off the first node (6V6 plates) so it is getting more like +285 V.

I swapped in a 12at7 into V1 but didn't hear an obvious increase in output volume with a test tone, and the tremolo intensity dropped way down.

I changed the tremolo injection point per your suggestion, connecting the CT of the PI trans sec. to the top of the Intensity pot (and lifting the connection from same to the V1b cathode circuit). The available tremolo intensity was reduced to almost nothing, so I undid the change.

With the 12AX7 the cathodes are at 1.3v for V1a and 2v for V1b.
With the 12AT7 the cathodes are at 2v for V1a and 4v for V1b.

Looking at the load lines, assuming the resistance of the Rp||PIT primary to be ~11k5, the load line for V1b is almost vertical. Does that seem right? I could use a lower Rk for that triode to bias the stage a bit hotter.

[imaradiostar]

I'm sorry about that. I used Merlin's load line plotter to guestimate the gain of each. I guess I wasn't estimating impedances correctly.

Are you using the champ front end now? What era champ? Are you using the lower cathode and plate resistors on the second stage into the transformer?

Not sure about the trem thing. I know there are amps out there that are set up that way- check out this Gibson schematic:

http://schematicheaven.com/gibsonamps/ga15rvt.pdf

Come to think of it- it probably needs a coupling cap and a resistor so it can pull the grids below their normal bias point. I dunno why I hadn't thought of

You're right on the load lines, assuming you're referring to a 12ax7. I'm kinda surprised the 12at7 didn't make a bigger difference.

I had a MMB amp and it blew up during practice at church- it was pretty awesome- crazy sound and smoke and everything. I built a mini bassman circuit in the chassis using 6bm8's and I really like it. That amp belonged to a friend so I gave it back to him once it was in good shape. I still have the interstage transformer and the eyelet board from the MMB amp and power and output transformers that are close enough for jazz so I think I might have to build something along these lines to play with- maybe I can better compare notes then!

I really want to play with the whole triode/pentode thing- I have a bunch of them and I think they sound pretty good. Maybe if I'm lucky I'll have a chance to try this idea with one soon.

jamie

[Groover]

The amp has a BF/SF preamp and vol/tone stack, with an added bright switch and a 25k midrange pot added.

I haven't changed the cathode resistor on V1b yet. I figure 1k might bias the 12AX7 at about -1.3Vg which would be a little hotter. I'll let you know how that works out.

How would you add the coupling cap and resistor to the trem injection point?

Once again I was only listening to a test tone, while swapping the 12AX7 and the 12AT7, but whatever difference there was in gain/volume wasn't very noticeable.

[Ronsonic]

I know this has nothing to do with your question, just want to point out that if you rewire the preamp section of the MMB like a Tweed Vibrolux or such it turns into a happy little screamer of a guitar amp.

[Groover]

I did initially rewire my MMB with a Tweed Harvard front end like yours, and it did sound very good.

Been there done that!

[imaradiostar]

The amp has a BF/SF preamp and vol/tone stack, with an added bright switch and a 25k midrange pot added.

I haven't changed the cathode resistor on V1b yet. I figure 1k might bias the 12AX7 at about -1.3Vg which would be a little hotter. I'll let you know how that works out.

How would you add the coupling cap and resistor to the trem injection point?

Once again I was only listening to a test tone, while swapping the 12AX7 and the 12AT7, but whatever difference there was in gain/volume wasn't very noticeable.

Take a look at that Gibson schematic for ideas. I guess what I was saying is there has to be an amount of added resistance between the center tap and ground and a capacitor to swing that point up and down which, in turn, alters the bias point of the output tubes. Because the trem signal is in phase at the output tube grids and the audio signal is out of phase, the trem signal acts only to alter bias and doesn't appear in the audio output. The cap charges up so that the pulses on it push the center tap positive and negative relative to ground- like the trem in a Fender Princeton Reverb. I'd probably steal the values from that circuit- .1uf and 250k trem level. I'd wager you could use a lot of different values and it would work but Fender probably put some R&D into it and arrived at those values for a reason.

If you use the Princeton verb oscillator you'd free up half a tube for added gain as well- then you could use the au7 side of a 12dw7 to feed the transformer and the extra ax7 half as a gain stage in-between- possibly even put the cathode bypass cap on a footswitch for "gain boost" since you'd have plenty of gain after that!

The BF Fender TMB EQ is great for clean sounds but I wonder if maybe it isn't the right way to head for this sort of amp. I'd go for the aforementioned tweed circuit, the Dr. Z Ghia style midrange circuit (like a modified Gibson single mid control), or the Traynor/rt66 tone control. Between those different designs you're likely to finder lower loss, more pedal friendliness, and tones that are a better complement to the easily over driven (possibly even AB2) output section.

Once again, haven't built this amp with those parts but I've built the above tone circuits in other amps and found them useful. If you'd like I can post some schematics with examples.

jamie

[Groover]

OK, I got my 12DW7 and I am wondering how to draw the load line for the "12AU7" stage - what value to use as plate resistance?

The plate load is a 15K resistor in parallel with the PIT primary.

[tubeswell]

OK, I got my 12DW7 and I am wondering how to draw the load line for the "12AU7" stage - what value to use as plate resistance?

The plate load is a 15K resistor in parallel with the PIT primary.

Get a tube characteristic chart with 12AU7 curves

The HT(B+) voltage - measured at the point where the plate resistor connects to the power rail - is plotted on the horizontal axis. (say its 150V)

Then divide that HT voltage by the value of the plate resistor, and you get the amount of current to plot on the vertical axis. (e.g.; say the HT voltage at that point was 150V: therefore 150V/15,000k = 0.01A (10mA)

Then connect up the two plots with a straight line. There you go.