I'm tired of anemic guitar amps!

[scruffy]

Okay well, all that remains is to borrow a '57 from the studio and get you guys some sound clips.

I'll give you clean, crunch, lead gain, and serious overdrive. You'll swear it's four different amps. And then I'll show you what the reverb can do, it's a lot of fun.

In addition to the PI and the component values for the cathode follower, I'd add this as the third discovery: dialing the presence frequency into the reverb springs. At the end of the day, this is what you want from your amp, to get a particular sound, and a particular responsiveness. The more of that you can get, the better and more useful the amp. Reverb spring noise is kinda-sorta around the same frequency range as string noise from a guitar or bass. So if you can get your presence control to dial that range it'll have a reason to exist.

If you looked at the Ampeg schematic, you may have noticed the bridged T filter before the effects loop, and the control called "ultra-mid". You can get notches of 20 dB and more from these things. Gibson used them, so did Magnatone. To make the best use of an effect like presence, you want to dial it in to specific frequencies on your guitar, so you need extensive control of the mids. A tone stack is not enough.

As a guitar player, at minimum I want upper mids and lower mids. Fender's midrange control raises and lowers them at the same time, but I want to distinguish them. You can use the online tone stack calculator to see the dip caused by the midrange control in the Fender and Marshall stacks, it's anywhere from 400 hz to 1 khz depending on the slope resistor and etc. Lower mids are 400-500 hz, let's say below 600. Upper mids are 700 hz - 1 khz or so. It is very easy to design bridged T filters centered at 500 hz and 850 hz, with a bandwidth of about 100-200 hz and a variable depth.

In my amp I use a 250k midrange control in the up-front tone stack. Somewhere around 100k it starts giving you lots and lots of big fat juicy midrange. (It's a different way to get mids, instead of cutting lows and highs). In the extreme at 250k it will basically disconnect the tone stack and allow your full signal through. So the bridged T filters want to be between the tone stack and the gain stages. The logical place to put them would be at the output of stage 3, after the coupling cap has been selected.

Finally, in relation to the holy grail of single knob control over both gain and voicing, the best that can be done is a knob and a switch. There are voicing changes in several stages, to accomplish the high gain sound. A 4PDT toggle would do it, but no such thing exists on a pull out pot. The only other option is head in the direction of foot switch control and adjust the pathways with LDR's like Ampeg did.

 

[scruffy]

The long and short of it is a non-anemic 200 watt amp will set you back almost 2 grand just in parts cost. It's 500 bucks just for the case and chassis, and the cheapest 200 watt output transformer is the $400 Orange replacement made by Mercury Magnetics. (The really good Marshall Major one is like $550). Add in an Antek toroidal PT for another $100 and that's 1500 just for case and iron. A quad of KT-88's is another 250 bucks, and the rest of the tubes are another 100, which only leaves 100 bucks for resistors and capacitors (and knobs, and terminal strips).

If you get very lucky, you might be able to find a broken 200 watt tube amp somewhere, and strip it and cannibalize the iron. It would be worth a thousand bucks if you don't mind the cosmetics. But 200 watt tube amps are pretty rare to begin with. And a 200 watt amp will "barely" fit in a JCM-800 large box, and it'll be heavy as hell. Maybe go modular 50 watts at a time.

 

[scruffy]

Picture of amp back in case (without faceplate):

Remounting the chassis is a bit problematic, since all the inside nuts are gone. I'm thinking some #10 U-nuts, we'll have to see if they can handle the weight. Otherwise, a series of L brackets along the back of the chassis. The front needs reinforcement too. See that little hole to the left of the knobs between the 2nd and 3rd set? That's what holds the front of the amp in. Two of those, that's it. We're going to have to do better than that.

 

[scruffy]

So far so good. One very long shielded wire is needed for the stage 4 tone control, and the coupling cap in the reverb amp needs to be adjusted (there's a lot of bass in the old reverb tanks). And, I think I accidentally used a 25 uF cathode cap in the follower, there's too much bass running through there. (It should be 1 or 2.2 uF). Everything else works perfectly. Just a little tone shaping remaining.

(And I have to go to the auto parts store, it's the only place that has U nuts in stock locally, and they only have metric sizes, so I'm gonna try M6).

 

[scruffy]

This is another amp Scruffy built for a recording artist. The little green things are the pots. Each of those yellow caps is worth 5 bucks, and the two red ones are 12 bucks each. Amp kicks butt. It's been on 50 recordings already. It's getting a maintenance, new tubes, check voltages, adjust bias, etc

 

[scruffy]

I've been researching the history. The reason Fender went to 12at7 PI's is because the AX's were farting out at high volume.

The output tubes have a maximum value for the grid leak resistor. In the case of a KT-88 it's 100k for fixed bias. 500 volts into 100k is about 5 mA, and a 12ax7 isn't up to the job, it does about 1.5 mA max.

In the case of the Marshall Major they use 68k grid leaks for pairs of KT-88's, which requires "even more" current. Hence the driver. (Which by the way, the value of that cathode resistor should be 600 ohms, not 1.5k).

 

[scruffy]

Had a heck of a time tracking down an intermittent problem in Stage 1, turns out it was a bad pot. 2 volts AC on top, nothing on the wiper. It was a very annoying problem, it would only appear when I put the chassis in the case.

Couple of real world experiences - as a bass amp it works better with the tone stack up front. (I just disable the master tone stack, it doesn't add anything to the sound, it's a bit too subtle for bass).

The secret tone shaping recipe for bass is to add a bridged T notch filter ahead of (or after, but ahead of works best) the tone stack, centered at about 225 Hz. You don't need much cut, even 6 dB helps a lot. If you have room for an extra pot you can control the depth of the notch.

You can cut all of the bass out of the first two stages of the amp, by setting the cathode bypass caps and with the tone stack, but you get it back again from the cold clipper. The cold clipper is amazing for bass, it's great. Makes my 12" speaker at 20 watts sound like an Acoustic 370 at full volume.

In real life I'm getting about 100 mV out of either a guitar or bass when strummed across the strings, and about 4 v p-p maximum output from the first amp stage, so that would be a gain of about 40, not bad. The second stage starts overloading with the gain set at about 1/2, and it has the same gain as the first stage, so it'll try to take the 2v signal up to about 80 volts, which is at the very edge of the AC load line with 350 v supply and 220 v on the plate.

80 volts is "sufficient" to get a very noticeable effect from the cold clipper, even with a 4.7k cathode resistor. Because the cold clipper is set up to draw more current you may get a "slight" amount of gain out of the stage but it won't be much. Using Soldano's 39k value for the cathode resistor, the gain will be near unity. With 4.7k you'll get some gain because you're drawing about twice the current as an ordinary amp stage.

So you could in theory feed 100v AC into the cathode follower, which would be disastrous because it has a gain of 40-ish too. It's vital to put a crunch pot at the input to the cathode follower. And that combined with your large-value grid stopper takes off some highs and gives you a darker sound. (Which you can then shape with the master tone stack if you wish, or a subsequent stage with additional tone shaping).

For bass, I use the cold clipper at about 20%, and the cathode follower set clean, right at the point of authoritativeness where you're starting to get some bass boost.

 

[scruffy]

Oh - so I've been through the entire voltage chart in both amps, and based on that I'm going to revise my prescription for the PI.

With a 12at7 at 450 volts supply, use the Bassman circuit here, but with 47k and 53k (not 51k) load resistors, instead of the 82k/100k in the schematic.

https://schematicheaven.net/fenderamps/bassman_5f6a_schem.pdf

With the 47k/53k combination your gain will differ by only .04 between the two sides of the PI. I tried it, it works.

Other changes:

Use two 820 ohm resistors in parallel instead of the 470 ohm resistor in the schematic. This will give you 410 ohms which is exactly the right value.

Use 330k on the grids instead of 1M, value not critical, 390k will work so will 270k.

Use a 4.7k fixed value resistor instead of the 5k pot. Connect the .1 uF presence cap to the top of the resistor and put it in series with a 25k presence pot to ground.

My feedback resistor is about 300k connected to the 16 ohm transformer tap. In theory that makes my feedback only 1-2% or so, but I'm getting audible gain reduction and the presence control works fine.

So that's my magic PI that will easily drive a quad of kt90's - 12at7 at 450 volts with 47k and 53k load resistors, 410 ohms cathode bias, and a 15k tail split into 10k and 5k.

For this arrangement to work (well), the grid bias resistors in the output stage should be no smaller than 100k. This arrangement will put your plates at 205-ish volts, meaning about 300 volts with the 90 volt tail. With a 100k AC load per side the load line says the useful swing will be 80v to 320v, or almost 250v p-p !

To get the 100 volts needed to drive a pair or quad of kt-88's, about 8 volts signal is needed at the input to the PI.

 

[toobfreak]

Had a heck of a time tracking down an intermittent problem in Stage 1, turns out it was a bad pot.

Glad to hear you are still working on this thing! I think I checked in days ago to see if I had missed anything, I'm curious to see it finished, maybe even kinda hear it. But then, I've done a lot of prototyping in my life and understand there is a huge chasm between an operational or working prototype as proof of concept, and a /FINISHED/ prototype you are proud to show off to others.

• To be OPERATIONAL, you merely just need to plug it in and it makes noise but no smoke.
• But to be FINISHED, it must continually and repeatedly pass multiple inspections where each time: visually, audibly, electrically, you keep finding little cosmetic defects, electronic tweaks and adjustments, whose further refinement and need are now revealed, until after effecting all of them you don't immediately see/hear yet another and another and these tweaks start becoming less and less frequent in occurrence or notice until after doing the last one, you begin not seeing any other thing immediately reveal itself as needing refinement too shortly thereafter.

Sometimes this latter stage of development can go on for days, weeks, months, and even years. The "finished" stage is that point where you sit down and use the thing, look at it and check it out and are 100% happy with it as a "perfected" item and cannot think of a single further thing needing changed for the time being.

PS: Before I forget, I meant to add what an unusual thing to have crap out on you but a brand new potentiometer! It reminds me of a really cool troubleshooting feature I have with one Fluke meter where you can set it up as a "recorder" to keep a record of a given measurement over a period of time (minutes/days) then you can set up a variety of triggers. This way, you can walk away and it will record when and how some event happened while you were away, or act as another person telling you what is going on while you are at the other end across the room.

 

[scruffy]

Glad to hear you are still working on this thing! I think I checked in days ago to see if I had missed anything, I'm curious to see it finished, maybe even kinda hear it. But then, I've done a lot of prototyping in my life and understand there is a huge chasm between an operational or working prototype as proof of concept, and a /FINISHED/ prototype you are proud to show off to others.

• To be OPERATIONAL, you merely just need to plug it in and it makes noise but no smoke.
• But to be FINISHED, it must continually and repeatedly pass multiple inspections where each time: visually, audibly, electrically, you keep finding little cosmetic defects, electronic tweaks and adjustments, whose further refinement and need are now revealed, until after effecting all of them you don't immediately see/hear yet another and another and these tweaks start becoming less and less frequent in occurrence or notice until after doing the last one, you begin not seeing any other thing immediately reveal itself as needing refinement too shortly thereafter.

Sometimes this latter stage of development can go on for days, weeks, months, and even years. The "finished" stage is that point where you sit down and use the thing, look at it and check it out and are 100% happy with it as a "perfected" item and cannot think of a single further thing needing changed for the time being.

PS: Before I forget, I meant to add what an unusual thing to have crap out on you but a brand new potentiometer! It reminds me of a really cool troubleshooting feature I have with one Fluke meter where you can set it up as a "recorder" to keep a record of a given measurement over a period of time (minutes/days) then you can set up a variety of triggers. This way, you can walk away and it will record when and how some event happened while you were away, or act as another person telling you what is going on while you are at the other end across the room.

Still in search of the holy grail single knob control - I'm coming to the conclusion it's not possible. There are so many different forms of clean and crunch that switches are needed.

So far these are the observations:

1. The cathode cap can be either 1 uF or 25 uF, other options are not needed. But this is for each of 2 (in the practice amp) or 3 (in the mother amp) preamp stages - so a 4-position 3 pole switch should add 25 uF to stages { none, 3, 2, 1 } in that order. The rotary switch could be replaced with a 4PST toggle, in which case it's either all 1 uF or all 25 uF.

2. The switched coupling caps in stage 3 should be: .02, .01, .005, .005, .002, .001

3. Appropriate values for the cold clipper's cathode resistor are 39k, 22k, 10k, and 4.7k. Scruffy likes 4.7k but you have to drive the hell out of it to get the clipping effect.

4. The coupling cap between the cold clipper and the cathode follower could be reduced to .005, this relieves some of the mud in the bass range.

5. The coupling cap coming off the cathode follower could also be reduced, it doesn't need to be 1 uF. It could go as low as .1

There are THREE essential functions, each of which has a range.

A. Clean - this ranges from jazzy clean to sparkle clean. Most of the sparkle comes from your up-front tone stack.

B. Crunch - this range is in the border or transition area between clean and overdrive. The expectation is that when you play hard you'll get some grit, but when you play quietly it's mostly clean.

C. Gain - for metal riffs and shredding. Lots of midrange, very little bass.

"Clean" is sufficiently different that it could have its own switch. In addition to adding 25 uF to three cathodes, there are four points of voicing that can be selectively bypassed. So that's 7 versions of "clean". They're all useful.

The "gain" mode generally has little or no bass in it. The trick is to adjust the amount of bass just right for your guitar. You want thump but you don't want beehives/fuzzies in your overdrive. So for this mode you have the same bass adjustment points, you just start at the opposite end, and add bass as needed. The up-front tone stack is almost irrelevant in this mode, it doesn't help you very much. Most of the useful shaping is done by the master tone stack, "after" all the grit has been added.

The crunch mode (I call it the southern rock mode) is the most complicated because it generally makes use of both tone stacks. There are so many different tone shaping options in this mode, you can combine the stacks with the cathodes and grid voicings in any combination.

You could do toggle switches per stage, or rotary switches by function, or both. On the mother amp there is a rotary switch for the cathodes, on the practice amp one of the volume knobs pulls out. On the mother amp there are pull-outs per stage for voicing, in the practice amp there's only one pull-out so you get either clean voicing or gain voicing.

Then there's extra fun with the reverb, I haven't started on that yet.

 

[toobfreak]

Still in search of the holy grail single knob control - I'm coming to the conclusion it's not possible. There are so many different forms of clean and crunch that switches are needed.

Sounds like the amp is capable of far more versatility and change than you need and you need to narrow it down to just the 2-3 distinct sounds you really need and will use, and maybe can tie all changes together through just a single or a few switches.

 

[Fort Fun Indiana]

Sounds like the amp is capable of far more versatility and change than you need and you need to narrow it down to just the 2-3 distinct sounds you really need and will use, and maybe can tie all changes together through just a single or a few switches.

I tend to agree. A useful tube amp is one that can be used live, given most studio producers now will just send your direct signal through a model anyway. Unless you are a star with some pull.

And the live player doesn't want to haul around gear with a bunch of what they consider to be useless deadweight. Most players find 2 or 3 base amp tones they like and do the rest with pedals. Many,only ONE amp tone they like. See: the very popular Fender and Supro amps.

Now, personally, I would love to have such a versatile amp at home as a way to practice tone shaping and to search for a tone I love. But that's a luxury that most players don't or won't have money or time to afford themselves.

 

[scruffy]

Sounds like the amp is capable of far more versatility and change than you need and you need to narrow it down to just the 2-3 distinct sounds you really need and will use, and maybe can tie all changes together through just a single or a few switches.

Yes! Research says I can build a 200 watt one of these on a long Marshall (27") chassis, or a 100 watt version in a small box. Both chassis only have 6 knobs, 4 for the tone stack and 2 volume controls.

 

[toobfreak]

Yes! Research says I can build a 200 watt one of these on a long Marshall (27") chassis, or a 100 watt version in a small box. Both chassis only have 6 knobs, 4 for the tone stack and 2 volume controls.

That sounds reasonable. If you need, for a price, you can probably buy some good quality multi-ganged rotary switches, that way, when you turn a selector, you are affecting 2-4 different variables within the circuit at once.

But if you go that route, take your time and think it through for future repairs, troubleshooting and upgrades that you don't paint yourself into a nightmare of wiring and complexity.

But before you can do any of that, you must be sure of exactly where the meat in the circuitry ends and when you begin cutting into the fat.

 

[scruffy]

I tend to agree. A useful tube amp is one that can be used live, given most studio producers now will just send your direct signal through a model anyway. Unless you are a star with some pull.

And the live player doesn't want to haul around gear with a bunch of what they consider to be useless deadweight. Most players find 2 or 3 base amp tones they like and do the rest with pedals. Many,only ONE amp tone they like. See: the very popular Fender and Supro amps.

Now, personally, I would love to have such a versatile amp at home as a way to practice tone shaping and to search for a tone I love. But that's a luxury that most players don't or won't have money or time to afford themselves.

IMO the amp should be loud enough to be your personal monitor on stage.

 

[Fort Fun Indiana]

IMO the amp should be loud enough to be your personal monitor on stage.

I agree, I HATE guitars through monitors. Might as well just use a modeler and leave the amp at home.

 

[scruffy]

That sounds reasonable. If you need, for a price, you can probably buy some good quality multi-ganged rotary switches, that way, when you turn a selector, you are affecting 2-4 different variables within the circuit at once.

But if you go that route, take your time and think it through for future repairs, troubleshooting and upgrades that you don't paint yourself into a nightmare of wiring and complexity.

But before you can do any of that, you must be sure of exactly where the meat in the circuitry ends and when you begin cutting into the fat.

Yes. The problem is the amp is multi staged and the stages are in series.

For example, one essential function is "cut bass". In the first two (or three) gain stages you can bypass the cathode with 1 uF caps, which will start cutting below 150 Hz or so. That works good for shredding. But you need the clean too, and usually for clean you want the bass. So how much bass are you going to cut and at what frequencies? The player needs some control, and it can't be gotten from a tone stack.

The best solution is to use a rotary switch and let the player dial the number of stages to bypass which equates with the amount of cut. If you want to get clever you can do different frequencies for each stage. I want to use the amp for bass too, which is a slightly different cut, it's a notch centered at about 220 Hz - and you can use that notch with or without any other kinds of cuts or boosts.

You don't want to replace the bass control in the tone stack though, so you need somewhere to put the switch. The biggest switch I could find on a pot is 4pdt and it's a huge pot, has a little daughter PC board. There is plenty of room to the right and left of the controls in the wide body chassis, could easily add another tone stack and a couple more volume controls. Or, just 4 miniature toggles.

 

[scruffy]

Another bad pot. Of the same type, from the same manufacturer. "Delicate", is what they are.

I did exactly what we talked about, in terms of bypassing the cathodes of the first three preamp stages. It works beautifully. Very intuitive. You have a rotary switch called "bass boost", and as you dial it from left to right you keep adding more bass. Left is thin, right is fat. Easy peasy.

The tone stack is a killer, it's taking a 4 volt signal down to a couple hundred mV. Basically negating the gain from an entire preamp stage. In the mother amp there is an additional clean preamp stage after the up-front tone stack to make up the gain. The practice amp doesn't have an up front tone stack, and also doesn't have (or need) the third preamp stage.

 

[scruffy]

These amps are definitely not anemic!

Stage 2 preamp is outputting 80 volts AC into the tone stack. The first stage generates 4 volts, for a real life gain of 40. Then the signal gets cut in half by the tone shaping circuitry around the gain control, so the 4 volts becomes 2. Then another gain of 40 takes the 2 volts up to 80.

Audible distortion starts happening about halfway up the gain knob, so maybe around 1 volt at the wiper, which becomes a clean 40 volts after stage 2.

You'll lose about 90% of your signal in the tone stack, so 80 volts will become 8 and 40 volts will become 4. 4 volts is about the highest totally clean level that can be fed into stage 3. Where it will be boosted by 40 times again. So it will recover from the brutal cut inflicted on it by the tone stack.

In real life you wouldn't set any of these controls more than halfway up. And the choice of coupling cap can have a significant effect on level too. I'll keep sharing measurements as I get them.

 

[toobfreak]

These amps are definitely not anemic!

In real life you wouldn't set any of these controls more than halfway up.

Still, I wonder if it might be worthwhile to cut the first stage to 2 or 3 volts? Reason? I rather like a device which utilizes the full range of its controls; rather than have half the range of the pot "dead room" where nothing good happens, being able to utilize most or all of the control makes the adjustment more sensitive by stretching the adjustment range out. And it might make some of the circuit happier not being driven so hard.

And the choice of coupling cap can have a significant effect on level too. I'll keep sharing measurements as I get them.

Do you mean value of capacitance or merely brand or type of capacitor (such as polystyrene vs. polypropylene)? Is there no way you can direct couple any of these stages or do none of the biases line up?