I'm tired of anemic guitar amps!

[toobfreak]

Okay, the front panel will show you why this amp needed real estate.

Wow, looking real nice, it has more knobs and switches on it than I realized!

 

[scruffy]

In the first pic above, the amp is on its left side, so the right side of the front panel is on top. The input Jack is on the top right. Immediately to the left (below, in the pic) of the input Jack is the rotary switch for the cathode bypass capacitors.

You can see how the knobs are arranged in columns of 3. So above the input Jack and the cathode switch, those two columns are your up-front tone stack. That's stages 1 and 2, and they function together as a unit but they're segregated by function for a very good reason. Stage 1 gives you only cathode control and a gain knob, nothing else. Stage 2 gives you the tone stack. You can bypass the tone stack, but not stage 2. If you bypass the up-front tone stack then stage 2 will run "raw", so you'll have two gain stages with a level control in between, like a high-gain Marshall.

Stage 1 gain is about 80, it'll boost your 5 mV guitar signal to about 400 mV. After stage 1 the signal level is cut in half by the ancillary tone shaping components that live around the gain control. That 200 mV signal is then applied to stage 2, so at maximum gain your stage 2 output will be 16 volts. Then once again the signal level is cut in half by the tone stack. So the maximum available signal level at the entry to the gain stages is about 8 volts. That's "max", in reality to retain a clean signal the gain control might be up around halfway, so we can use a nominal 4 volts.

If you bypass everything and apply that 4 volts directly to the power amp, you'll have only a master volume in between, and it will give you full output at a setting of "2" unless you switch in the master tone stack, in which case you can go to "4". Everything above 2 or 4 represents power amp overdrive.

Or, you can go the other way and back off on the preamp gain to get a cleaner more controllable sound. The headroom is there for a reason - generally though, the level settings will be down around 2 or 3 in all the stages for most normal purposes, which will cover about 95% of rock n roll.

So the front panel goes right to left in columns of 3 knobs each. The first two columns are stages 1 & 2 and they're always in. Then, from right to left there are stages 3, 4, 5, and the reverb, all of which are switchable. Followed by the master section, which consumes two front panel columns (because it has a tone stack) and is always engaged (although the tone stack can be bypassed, just like the up-front tone stack).

The original chassis had two switches and two knobs directly in front of the power transformer, those are removed and I'll put a plate over the holes. The missing control on the front panel is a 250k pot for the reverb mixer (I ran out), it should arrive shortly.

 

[scruffy]

Better pics:

 

[scruffy]

The two loose wires are the reverb send and return.

 

[toobfreak]

Better pics:

Looks killer! My only thought about the front panel is that I might have tried to slightly group the knobs and switches according to function and relationship, but you might have actually done that and it just doesn't show yet or maybe you'll do it later with lines or markings of some kind on the front panel.

My other thought is that I trust the front panel will have some kind of power light to indicate it is on.

Lastly--- there is what appears to be a 1/4" jack right next to the PT--- what's that for?

 

[scruffy]

In traditional guitar amps, the most important design point is the transition from clean to breakup. Players want that to be seamless and smooth.

And, it's easy to kill this transition with a poor amp design. Even something simple like over filtering the power supply can affect it.

The early amp designers tried everything in terms of designing around this transition. For example check out the Park 1210 schematic, they use two cold clippers in a row with the second being more severe than the first. So as you raise the gain you get grit, then more grit, then overdrive. The player simply sets the preamp gain to match the range of the guitar's volume knob, they're very easy to use and they work well for that purpose.

But then you have the Marshall Major, which is notorious for lighting on fire and blowing up. The problem is the voltages are too high for a measly 6550 tube. A KT-88 would do just fine though, as it can handle 600 volt screens (instead of 440).

The original AIMS amp, may not have worked at all with modern 6550's. And if you drop in KT-88's instead you have to bring the bias up another 20 volts. In this and Fender designs (AIMS is basically a Fender design, just different output tubes), the PI is responsible for most of the grit. The same is true for a Marshall but it starts earlier. However the Major is a different beast, it's like a Princeton on steroids. There are two common reasons for arcing over: one is, people forget to plug in the speaker, and two is, people put pedals in front of the amp with lots of harmonics and distortion. The old story is true, about the FuzzFace destroying the Marshall Major.

There are mods you can do to alleviate these conditions, the most important of which are increasing the size of the screen resistor and installing protection diodes around the output transformer. These mods only change the sound and responsiveness "a little", not enough to matter. To get the KT-88's to start clipping you have to exceed the plate voltage from the power supply. In my case that's 530 volts, and 100 watts is 600 volts across the OT primary, so between 530 and 600 I'm clipping. That's maybe 90 watts and above. And the maximum amount of clipping is very small before full power. So in this amp, there really is no transition "in the output stage", it occurs elsewhere.

If you push the amp input to 10 or 20 volts, what are you really pushing? Well, first the PI, then the driver. The KT-88's require a large input signal to get to full power, maybe 100 volts or so, which is about the limit of what the driver can provide. So most of the crunch is going to come from the PI. The difference between a clean 0.8 volt input, and 100/6 / 10, is significant. So basically the control over responsiveness "ends" at the PI, whích is definitely a different design philosophy from Fender's tweed era

 

[scruffy]

Antek is now selling toroidal OT's!

100 watts for 120 bucks.

And, they make 5H chokes that weigh less than a pound. You can do all the iron on a 100 watt amp for less than 20 pounds.

 

[scruffy]

Here's the beginning of the schematic, which I'll share with you. Very simple stuff, very straightforward.

The amp takes 312+40 mA at full power. The full wave rec allows the bias tap to be used. The bias control goes down to the high 60's so either 6550 or KT-88 tubes can be used.

Loaded B+ is 535 volts, that's what goes on the plates and screens of the output tubes. The general rule is two stages at a time without decoupling. The 450 volts is for the PI and the driver, both 12au7's with 47k-ish loads. The 400 volts is for the reverb transformer. The 375 is for the two cathode followers. 360 is for the cold clipper and its driver. And 350 is for stages 1 & 2.

The fuses are large because of inrush currents at power on. The toroidal PT has its own inrush current and the many large capacitors take quite a bit of juice to charge up. The resistors in the bias circuit are small because we need every last bias volt for some of the newer production KT-88's. JJ's are currently installed, they're pretty friendly, they're coming in at -88 volts for the recommended 50 mA biasing (per tube). I'm also using the JJ filter caps, I've found them to be very reliable after at least 15 years in the field (being played every night).

We already went through the calculations for the resistors, not much to add there. The first resistor drops about 40 mA, I figure that's about 5 mA each for four 12au7 sections, another 10 mA for the class A reverb driver (two 12at7 sections in parallel), and 10 mA for everything else (ax7's).

Now we get to find out if the amp behaves. Two considerations are: 1. there is LOTS of gain, so decoupling through the power supply, and 2. negative feedback now encompasses three stages instead of just 2, so phase as the volume changes, and proper behavior of the presence control. All the level controls have to rotate through their ranges without any howling.

The current condition is, the amp passes audio at low volumes, and all stages except the reverb work. I had to take the tank out again because the stupid terminal strips are too tall, they have to be upside down to fit with the tank mounting. I'll leave this issue for next week and focus on the power supply for now. If there's no howling we can move on and see if we can get any from the reverb.

This amp LOVES big speakers, it sounds totally massive through an EVM-15L. I'll have to drag some gear to the studio and crank it and see what happens. The acid test is, I have a pedal called Cold Shot (made in Brazil) that's an amp killer, if my amp can survive that at full volume we're home free. The pedal can generate almost pure square waves at humongous levels, which very much taxes the output tubes and the OT. If you look at the plate signal at 100 watts you can see big huge bumps on the rising and falling edges of the square waves, that's the OT demagnetizing and generating more than a kilovolt of reverse voltage across the tubes (that's what the protection diodes are for, if the reverse voltage ever causes arcing in the tubes or inside the OT you're screwed, without those diodes you'll lose the OT and both tubes).

But with KT-88's we're well within spec on all fronts, we should be able to overdrive the output section all day without any complaints (except from the neighbors).

 

[toobfreak]

Here's the beginning of the schematic, which I'll share with you. Very simple stuff, very straightforward.

Effective analog is always deceptively simple, but really mostly because good analog depends on each device filling several roles and working energetically with other parts for various combined desired effects very much like a musical instrument does working together as one.

In that sense, a good analog circuit is "tuned" electronically to perform along a desired path much as a room might be tuned acoustically to behave properly.

 

[scruffy]

I'm becoming a big fan of 12au7 tubes. They sound great, and behave nicely. The cost is slightly higher current.

The 12at7 is nice in the PI position, and my best advice is DON'T use that god-awful 100 ohm resistor in the Fender amps. Instead, use the exact same AX circuit from the Bassman 5f6a, but use 51k/47k load resistors with an AT. This will give you tons and tons of clean headroom, AND allow your presence control to work in the normal way.

Where the AU really shines is as a replacement for the Fender/Marshall cathode follower. That's a poorly designed circuit in the first place, the CF actually ends up stealing current from the preceding gain stage. The quick and dirty fix is to change the load resistors to 220k (instead of 100k), but a better solution is to use an AU. With an AU you get 52 volts input before breakup.

 

[scruffy]

So here's a question for you guitar players:

Would you rather use pedals to get crunch, or would you rather have the amp do the dirty work for you?

 

[Fort Fun Indiana]

So here's a question for you guitar players:

Would you rather use pedals to get crunch, or would you rather have the amp do the dirty work for you?

I like having several gain stages on an amp. I really like my two amps that have 4 channels each (Soldano and Highes Kettner).

 

[sparky]

So here's a question for you guitar players:

Would you rather use pedals to get crunch, or would you rather have the amp do the dirty work for you?

I'm unsure how to answer that Scruff .......i've played with a LOT of guitarists , the tenured ones always cultivate their 'sound' ........i will say it appears many go through a pedal phase , only to forgo them and focus on their amps......thus the gibby/marshal and the strat/fender twin camps.....~S~

 

[scruffy]

I'm unsure how to answer that Scruff .......i've played with a LOT of guitarists , the tenured ones always cultivate their 'sound' ........i will say it appears many go through a pedal phase , only to forgo them and focus on their amps......thus the gibby/marshal and the strat/fender twin camps.....~S~

Yes. I ask because it's technically difficult to build the perfect crunch amp and the perfect clean amp in the same box. The differences transcend voicing and channel switching. The behavior of the PI is a prime example. In a crunchy amp you want breakup in the PI, it should start breaking up well before the output tubes. In a clean amp you want the PI to stay clean till we'll beyond the point of output tube breakup.

Expanding on the previous example, a 12at7 PI has a gain (per side) of about 13, so if you need 100 volts to drive the output tubes to full power you'll need maybe 8v at the input to the PI. The PI itself will swing twice as much voltage before reaching the endpoints of the load line, so you have all the headroom between 8v and 16v input where you get nothing "but" output tube saturation. What you "hear" through that range is all output tubes, then if you keep raising the volume eventually you'll hear the PI kick in with its own distinctly different flavor of crunch.

You can play with this ratio, if you like the PI crunch you can get it earlier by adjusting the load resistors (changing the gain). In Marshalls and many other amps they use a 12ax7 PI which breaks up a lot earlier, say around 4 volts (because it has higher gain, in the mid-20's). With this tube you'll get PI breakup long before the output tubes saturate, so as you're raising the volume that's what you'll hear first.

In this particular amp I've changed the rules of the game, taken all the crunch and moved it up front. The driver tubes will deliver 250 clean volts to the KT-88's. That's enough to grow the signal 2.5 times beyond the point of maximum power. That's some sweet output tube compression, and this is how you run in Tweed mode. You push the clean signal up to 250 volts and it stays clean all the way through, until the very end.

For dirt, the reality is the sweet output tube overdrive is going to get completely lost in the mix of all the other stuff that's going on. The shredders I know like to layer on multiple crunch tones to get the midrange they need, if you look at their front panel all the knobs are like 7 or 8, they're using "all" the available forms of crunch. That signal is going to be dirty as hell before it ever touches the PI, with a gritty PI it'll get even dirtier, and a little additional output tube crunch isn't going to make a difference one way or the other.

That's the philosophy I used in this design. Super clean power amp, tunable for thump and presence. Detailed control over up-front crunch character, and if you don't want crunch simply bypass the stages. And this is why the early stages of this amp are segregated by function. If you DO want crunch then your early sound has to be significantly shaped (cut bass, boost mids, etc), and this shaping occurs in a characteristic order which is neither needed nor wanted in a clean amp. If you DON'T want crunch then you remove this shaping and you can literally run "straight through" with only level controls in the way and with all your tonal variation coming "only" from your guitar pickups and guitar knobs. That's as pure as it gets, there's nothing cleaner.

For crunch, I put the voicing "in" the stages, and what that does is give the player the required voicings as the guitar's volume knob is being raised through its range. If you're clean at level 2 you want to hear the lows and highs, whereas if you're shredding at level 9 you want more mids and less extremes. So each gain stage has its own voicing and its own tone control, and the responsiveness of the instrument depends on when each stage kicks in and how it's programmed.

The idea is, the player shouldn't have to futz with (foot) switches while performing. Set up the amp, then forget about it. Focus on the guitar, and learn to control the volume knob while you're playing. You can do as much with the volume knob as most people can do with pedals.

I think you'll find, this approach is superior to the old school "crunchy amp". It avoids having to take the amp too far into the pigeonhole of crunchiness (beyond the point of no return). The clean sound has to be full and rich, and the thump has to be full and rich in a different way. We can accommodate them both with a good design. We'll see how this prototype behaves in the field.

 

[scruffy]

So here's a warm up shot of the practice amp, it's on a JCM-800 chassis with the old iron, which just happens to fit quite well. It'll go into a small box head.

This will be a non-ultralinear 100 watt KT-88 "classic" amp with a 12at7 PI and it'll need a choke (a JCM-800 choke will do fine, 5 H at 120 mA).

It'll also need an extra filter can because I'm going to add a tube and I don't want to run the PI at 600 volts. Yikes!

So I have to punch two holes in a steel chassis, that'll be fun. But the rest of it is easy. I can even use the Marshall faceplates, they fit perfectly and all the labels are useful. I'm looking for a place to put the bias pot, it'll go on the back panel but I'd rather not expose it there.

On this amp we're going to need 8v at the input of the PI to get to full power. There are two input jacks, so one will be clean and the other will be dirty. There will be a master volume to handle level changes.

There's not much room in this chassis to get fancy, I'll have to mount the remaining parts to find out. We could do things like concentric pots and switched pots and whatever, if any of it fits. So far there's no knob problem, the obligatory preamp gain control takes up one volume slot and the master volume takes up the other, and the tone stack is exactly what it is. If we wanted to add a resonance control we could put a concentric pot in the "presence" position and it would still make sense.

Anyway, this one might get done before the other one does.

On this one I'm going to try to fix the cathode follower by increasing the load resistance. While carefully monitoring cathode voltage to stay within spec. It should in theory be possible to get a clean sound through the cathode follower.

 

[Chuz Life]

So here's a question for you guitar players:

Would you rather use pedals to get crunch, or would you rather have the amp do the dirty work for you?

I have both pedals and a (Marshall Valve state) amp.

Out of laziness, I just use the amp for crunch. If I were playing on a stage, I would probably go for the pedals.

 

[Chuz Life]

So here's a warm up shot of the practice amp, it's on a JCM-800 chassis with the old iron, which just happens to fit quite well. It'll go into a small box head.

View attachment 1169872

This will be a non-ultralinear 100 watt KT-88 "classic" amp with a 12at7 PI and it'll need a choke (a JCM-800 choke will do fine, 5 H at 120 mA).

It'll also need an extra filter can because I'm going to add a tube and I don't want to run the PI at 600 volts. Yikes!

So I have to punch two holes in a steel chassis, that'll be fun. But the rest of it is easy. I can even use the Marshall faceplates, they fit perfectly and all the labels are useful. I'm looking for a place to put the bias pot, it'll go on the back panel but I'd rather not expose it there.

On this amp we're going to need 8v at the input of the PI to get to full power. There are two input jacks, so one will be clean and the other will be dirty. There will be a master volume to handle level changes.

There's not much room in this chassis to get fancy, I'll have to mount the remaining parts to find out. We could do things like concentric pots and switched pots and whatever, if any of it fits. So far there's no knob problem, the obligatory preamp gain control takes up one volume slot and the master volume takes up the other, and the tone stack is exactly what it is. If we wanted to add a resonance control we could put a concentric pot in the "presence" position and it would still make sense.

Anyway, this one might get done before the other one does.

On this one I'm going to try to fix the cathode follower by increasing the load resistance. While carefully monitoring cathode voltage to stay within spec. It should in theory be possible to get a clean sound through the cathode follower.

Dammit scruffy, you have been shopping for shit I don't even have time for! LOL

 

[scruffy]

Dammit scruffy, you have been shopping for shit I don't even have time for! LOL

View attachment 1169901

Yes, I've been shopping online, I got mine from MojoTone because it doesn't have that nasty rectangular hole for the PT. And, it has a second large hole on the other side of the choke for an additional LCR. Unfortunately it's a steel chassis, but that was the compromise.

It'll work fine though. Here it is mounted in the case, you can see the PT barely fits inside and the tubes are about that tall give or take.

I'd like to fix the cathode follower on this amp. One idea is to use a bootstrap design that raises its overall gain to about 80, while reducing distortion to near-zero. Another idea is just to increase the load resistor, the problem there is the heater-cathode voltage is already on the hairy edge of spec, the cathode is near 200 volts in the standard circuit. A third idea is to use a 12au7 which won't start distorting till about 5v drive, so if you cap the drive at 8v then 5/8 of your gain range will be completely clean and there will only be "a little" grit in the remaining 3/8.

The gain of the cathode follower itself could be just about as high as the mu of the tube. I'm going to use my patented two stage preamp, which takes a 5 mV guitar signal to 16 volts or so, and it can be easily cut to 8. With one gain stage only, the input to the cathode follower is no higher than 0.8 volts, therefore if the C/F itself has a gain of 80 its output is still 64 volts, which the tone stack will then cut back down again.

Here's the front of the amp. You'll notice the labels are correct. Based on the above math we would like an additional gain of about 10 when we plug into the "high sensitivity" jack. This will take us over the threshold for generating distortion in the cathode follower. Otherwise, the additional gain is not needed because the C/F will put exactly the maximum clean voltage on the input of the PI.

We can also run this without the cathode follower, if you plug into the high sensitivity input you get a very clean signal at the preamp output. And, since the high sensitivity channel has the tone shaping for gain, you get less bass and more highs.

With all stages in you can overdrive the cathode follower by about 100%, which is crunchy but not searing. With the cathode follower out and plugged into the low sensitivity input you only get a clean 10 watts, so practice volume. That's because the single gain stage only puts a maximum of 0.8 volts on the input of the PI, which is about a tenth of full power.

 

[Chuz Life]

Yes, I've been shopping online, I got mine from MojoTone because it doesn't have that nasty rectangular hole for the PT. And, it has a second large hole on the other side of the choke for an additional LCR. Unfortunately it's a steel chassis, but that was the compromise.

It'll work fine though. Here it is mounted in the case, you can see the PT barely fits inside and the tubes are about that tall give or take.

View attachment 1170188

I'd like to fix the cathode follower on this amp. One idea is to use a bootstrap design that raises its overall gain to about 80, while reducing distortion to near-zero. Another idea is just to increase the load resistor, the problem there is the heater-cathode voltage is already on the hairy edge of spec, the cathode is near 200 volts in the standard circuit. A third idea is to use a 12au7 which won't start distorting till about 5v drive, so if you cap the drive at 8v then 5/8 of your gain range will be completely clean and there will only be "a little" grit in the remaining 3/8.

The gain of the cathode follower itself could be just about as high as the mu of the tube. I'm going to use my patented two stage preamp, which takes a 5 mV guitar signal to 16 volts or so, and it can be easily cut to 8. With one gain stage only, the input to the cathode follower is no higher than 0.8 volts, therefore if the C/F itself has a gain of 80 its output is still 64 volts, which the tone stack will then cut back down again.

Here's the front of the amp. You'll notice the labels are correct. Based on the above math we would like an additional gain of about 10 when we plug into the "high sensitivity" jack. This will take us over the threshold for generating distortion in the cathode follower. Otherwise, the additional gain is not needed because the C/F will put exactly the maximum clean voltage on the input of the PI.

View attachment 1170197

We can also run this without the cathode follower, if you plug into the high sensitivity input you get a very clean signal at the preamp output. And, since the high sensitivity channel has the tone shaping for gain, you get less bass and more highs.

With all stages in you can overdrive the cathode follower by about 100%, which is crunchy but not searing. With the cathode follower out and plugged into the low sensitivity input you only get a clean 10 watts, so practice volume. That's because the single gain stage only puts a maximum of 0.8 volts on the input of the PI, which is about a tenth of full power.

LoL.

My typos are driving me ******* nuts.

I was trying to say "damn you, you have me shopping for shit I don't have any business shopping for"

It doesnt cost much to look though, so it's all good.

 

[Mr Natural]

Nice job!