[rob]

I think we need you to build us one, Brad!!!

Rob

[SarnoMusicSolutions]

Man, I appreciate the nudge.

I just got back from watching a gig here in town where a guy is using the SMS Classic pre into a Fryette 2/90/2 (formerly VHT) tube power amp. It's about 90 watts per channel, and each channel is driving a single E120 in a 2-12" Hard Truckers type cab. He's got the 2-cord fx loop, a buffer, a DiMarzio Dual Sound split to single coil and in the middle pickup position, etc. It's a very Jerry kind of rig. It sounds amazing, and really I'd have to say that the Fryette with the SMS Classic is hard to top. But it's a tube power amp, so it's gonna get that earlier '70s, pre-Mc2300 Jerry kind of clipping and dirt if driven that hard. The thing I think that's so special about the Mc2300 is that it's a transistor amp that you can bang on, abuse, clip, and just lay into like it's a tube amp. But the Mac can handle it if you're within reason. Get the Mac to peak, but don't go too far past that, and your guitar notes, especially the harder picked stuff will live in this region where the ice-picky peak transients are nicely smoothed out so they don't hurt, but instead they kind of squirt and with a bit of distortion on it, but still mostly a squeaky clean tone character. That's what I hear the Mac doing, it's dynamically behaving like a tube amp being overdriven a bit, but tonally, harmonically, it's far cleaner since the dynamics are being ridden by a limiter. On a tube power amp, once you've passed the peak threshold, it' just gets more and more dirty.

I think many of us have discovered that if you play a clean tube preamp into a solid state power amp and you have tons of clean power headroom, the tone and the dynamic "feel" is just too clean an sterile, no sustain, no smoothing of the pick attack, no squirt or juicyness or sting, just clean, wimpy, and kind of boring.

So back to the topic here, I'd love to develop a limiter device to behave like a peaked out Mc2300 but in a small package. I've got some ideas and topologies drawn up conceptually, but nothing physical yet. I do intend on hammering this out because I think this whole Jerry rig and McIntosh magic is something special and very unique and potentially very useful to many guitar players of many styles. And also, maybe more than anything, I find great personal pleasure in trying to get my guitar to sound like Jerry's did, at least sometimes. That tone, that sound just gets me. So if this thing can pull off the right character of peak limiting as well as create very similar subtle distortion overtones and the same smoothing and sweetening that the McIntosh's output iron offers, it could become a very cool addition to some of our preamp based guitar rigs.

Brad

[bcresci]

I've got some ideas and topologies drawn up conceptually, but nothing physical yet. I do intend on hammering this out because I think this whole Jerry rig and McIntosh magic is something special and very unique and potentially very useful to many guitar players of many styles. .. it could become a very cool addition to some of our preamp based guitar rigs.

Brad

Consider me a volunteer for the beta program.

[RiverRat]

So back to the topic here, I'd love to develop a limiter device to behave like a peaked out Mc2300 but in a small package. I've got some ideas and topologies drawn up conceptually, but nothing physical yet.

I finally did some research on the MC2300 Limiter circuit...

The One problem I see with designing the limiter as a standalone circuit is the limiter placement in the signal chain of the 2300... It sits between in the input circuit and the output transistors. Integrating it into the preamp and the end of it's signal chain now puts it in front of the input circuit of a power amp.

The limiter circuit isn't that detailed: 8 resistors, 2 caps, 2 diodes and 2 transistors. It's how it interacts with the output stage that's the magic.

[SarnoMusicSolutions]

Exactly. It actually is keyed or triggered into action by measuring the voltage across the output transistor output resistors, and that voltage essentially reveals the amount of current flowing thru the output devices. That limiter circuit is still a local feedback circuit that lives in the driver section, it doesn't go all the way back to the input driver board section. But either way, I agree that what's critical is that the limiter reacts to real output transistor power activity. So that's why if I do develop a device to emulate this "effect", I'd want to create a very scaled down power amp type circuit that also has some kind of transformer or output iron as well. Basically doing this power/current limiter thing but on a very small scale and at line level. Sure it won't be exactly the same as what's really happening at the Mac's output with all that high current and power, but the idea is to get close and with a practical device that you could theoretically operate at any level of loudness. Emulating the timing, the ratio, and the harmonic dirt that results would be the primary goals. I'm sure nothing will exactly do what the real thing sounds like. Also, Healy said that it's likely that another factor in Jerry's sound is not only the McIntosh being pushed, but also the speakers themselves being pushed real hard, magnets and coils beginning to saturate and even perhaps distort a bit too. All that alchemy will be hard to exactly duplicate, but maybe we can get something so close that it's good enough for most of us.

What's also worth noting is that this type of limiter in the Mac is called a VI limiter and can be found in nearly ANY power amplifier made since the '70s. Every amp designer will likely calibrate it and tweak it differently, but the same idea all around. What makes the Mac a bit different is that the output is loaded by the autoformer, and that involves inductive loading and iron core saturation and other factors that make the Mac very unique and extremely rugged and robust. But this over-current VI peak/protection limiter can be seen almost universally in transistor power amps of all types and makes. Maybe Mac came up with it, I'm not sure. What they did patent eventually was the Power Guard opto limiter circuit, which is different. Not sure if they claim any design patents on this Sentry type limiter.

Brad

[RiverRat]

What makes the Mac a bit different is that the output is loaded by the autoformer, and that involves inductive loading and iron core saturation and other factors that make the Mac very unique and extremely rugged and robust.

This should actually be "The source output is loaded by the reflected load and the autoformer in parallel"
or
"reflected load in parallel with the inductive reactance of the autoformer"

This topic reminded me of this thread:
http://www.rukind.com/forum/viewtopic.p ... =impedance

One good quote from that thread:
"But when I get the volume up so that the clip lights are firing when I hit my loud notes, thats when mcintosh really shines."

I had quoted some stuff from Roger Russell's website regarding the history of the autoformer and sentry monitor, but I had neglected to see how the two work in tandem with the output transistors... Now it becomes clearer!

I'm glad I went back and reviewed that thread, Roger stated on his site that DC isolation is not an issue with the autoformer, while others say that the autoformer in conjunction with a 1:1 transformer for DC isolation is a good practice in terms of safety.

So that's why if I do develop a device to emulate this "effect", I'd want to create a very scaled down power amp type circuit that also has some kind of transformer or output iron as well. Basically doing this power/current limiter thing but on a very small scale and at line level.

The autoformer route should be possible at line levels, combining it with a VI limiter and some low power transistors shouldn't be that difficult.

Code: Select all

Inductance gives you a impedance = 2*pi*Frequncy*L(inductance) or Z=2piFL. Its this simple formula that tells you what you need to know about the impedance presented ot the source.

Ripped straight from here:

http://www.intactaudio.com/atten_FAQ.html

Whether it would perform like the original is unknown... Thermal distortion within the output transistors could also be a factor in the original recipe!

[tigerstrat]

Also, Healy said that it's likely that another factor in Jerry's sound is not only the McIntosh being pushed, but also the speakers themselves being pushed real hard, magnets and coils beginning to saturate and even perhaps distort a bit too. All that alchemy will be hard to exactly duplicate, but maybe we can get something so close that it's good enough for most of us.

Another benefit (besides portability) of using only a single E-120 in my rig, rather than a 2x, 3x, or even 4x- more pushing, less overall volume. I'd think this part of the equation much more audible than any limiter dirt. Either way, when you are in a situation requiring you to cut your stage volume considerably, you kinda lose both of these gritty nuances.

[playingdead]

Not so sure about that last part ... the testing I did in my studio close-miking low, loud and really f---ing loud volumes with my own E-120s, there was little or no difference to the recorded guitar sound on playback. I used an 800-watt power amp and had those cones rocking. It was impressive how much volume they are capable of producing without any breakup of the sound. It was so loud it was painful with earplugs in.

My money's on the Mac itself, not pushing the speakers with a lot of non-MC2300 watts.

[SarnoMusicSolutions]

Consider that for a very long span of time with the Mc2300, Jerry had K120's in there. The alnico K120 is far easier to saturate than the ballsy ceramic E120. Not sure when he made the general shift to E's, but I'd guess it was in the mid 80's. I agree, it's really freaking hard to saturate an E120, and if Jerry was running 225 watts into 3 speakers, or maybe 300 watts into 4 of them, that's 75 watts per speaker and won't likely stress out an E120 much. But that is pushing a K120. And strangely, he did blow E120's on occasion, so go figure. The power per speaker calculations are based on the notion that in the years after the Wall he mostly used only one 300-watt half of the 2300. I'd bet that he more often blew them during MXR distorted and loud solos. Throwing square waves at the amp maxed out is gonna send a pretty brutal signal to a speaker.


Brad

[playingdead]

Very true ... my K-110s have a marvelous creamy breakup to them that is very musical. They sound great loud.

There's a new company called RedWirez that does high-quality impulse response files of various speakers; they use 8 different mics (including a 421) at four different cone positions and a few different distances. People are raving about how good they sound via the various digital devices (ahem).

One of the vintage speakers in their "to come" list is a K-120, which I'll be eager to audition.

[SarnoMusicSolutions]

Yeah, bring it on; a K120 miked by a 421 at about 2" just at or outside the edge of the dustcap. That's the impulse response I want! I guess we can make them ourselves too.

Maybe Axe-FX can model a Mc2300 clipping/limiting?????



Brad

[tigerstrat]

Consider that for a very long span of time with the Mc2300, Jerry had K120's in there. The alnico K120 is far easier to saturate than the ballsy ceramic E120. Not sure when he made the general shift to E's, but I'd guess it was in the mid 80's. I agree, it's really freaking hard to saturate an E120, and if Jerry was running 225 watts into 3 speakers, or maybe 300 watts into 4 of them, that's 75 watts per speaker and won't likely stress out an E120 much. But that is pushing a K120. And strangely, he did blow E120's on occasion, so go figure. The power per speaker calculations are based on the notion that in the years after the Wall he mostly used only one 300-watt half of the 2300. I'd bet that he more often blew them during MXR distorted and loud solos. Throwing square waves at the amp maxed out is gonna send a pretty brutal signal to a speaker.


Brad

I get more of a sense or a hint of some slight speaker loading when running the octave fx and the Mutron.

[drewfx]

An impulse response will not capture the distortion characteristics of an overdriven speaker, only the frequency response and time based "reverb" response.

drewfx

[RiverRat]

Also, Healy said that it's likely that another factor in Jerry's sound is not only the McIntosh being pushed, but also the speakers themselves being pushed real hard, magnets and coils beginning to saturate and even perhaps distort a bit too. All that alchemy will be hard to exactly duplicate, but maybe we can get something so close that it's good enough for most of us.

Another benefit (besides portability) of using only a single E-120 in my rig, rather than a 2x, 3x, or even 4x- more pushing, less overall volume. I'd think this part of the equation much more audible than any limiter dirt. Either way, when you are in a situation requiring you to cut your stage volume considerably, you kinda lose both of these gritty nuances.

The autoformer negates some of this, it designed for attenuation and not merely impedance matching... changing from 4ohm to 2ohm cuts the resistance in half and therefore increase the amount of signal flowing to the speakers... in the other thread, I quote Roger Russell say that the MC2300's output stage is designed to see a 2.1 ohm load... The autoformer is there to make sure the output stage sees the proper load impedance as well as attenuate the output signal and feed it to the load... the limiter is there to make sure that if the load gets to low or the gain is too high, it shunts the excess current away from the transistors so they don't become little toasters and go POP!

The best analogy for the autoformer is that it acts like a gearbox in a car... in 1st gear , car with the engine running 3000rpm may only be doing 20mph... In 4th gear, that same engine @ 3000rpm may be doing 50mph! The velocity increased but the engine speed is still the same.

I get more of a sense or a hint of some slight speaker loading when running the octave fx and the Mutron.

The load does not change. You are hearing either an increase in gain, or some distortion in the source signal. The load is the speaker, the pedal is just part of the source chain. signal flow source to load, so loading is the reverse... You load the source.

[SarnoMusicSolutions]

According to Roger Russell, the autoformer is there to match impedances so that full power can be delivered to various loads. While the amp at its core (with no autoformer) may be designed to drive 2 ohms, the autoformer allows for full power matching to the other load options.

Roger said: "Full continuous amplifier power could be delivered to each of these loads."

Also, he describes a feature of the autoformer in its ability to shunt current thru its own low impedance which can help protect speakers, but that is a completely separate thing from the limiter circuit. The limiter is a feedback network with transistors that behave much like switches that will shunt the audio signal before it reaches the power transistors, and is activated by detecting high amounts of current flowing thru the output transistors.

Here's his reference to the autoformer's ability to shunt DC at the output:

"The autoformer also protected the speakers from damage in the event of amplifier failure. Should a direct current component appear at the amplifier output, it was shunted by the low DC resistance of the autoformer, instead of passing through the speaker voice coil, which could damage the speaker or even cause a fire..."

Brad