Help me design my first amp!

TheWinterSnow

Den Mørke Natt
Oct 22, 2008
3,087
4
38
34
Sacramento, CA
I have been researching the past couple of weeks trying to design my first amp, including adding some of my own ideas and taking from other amps. I have been trying to make an amp that has just about every feature in the book that most of us would want in an amp with the type of tone that we all enjoy while dealing with the flaws that have been apparent in a few of the favs.

I am trying to pick up a consensus regarding mostly two amps, the beloved 5150 and 2ch Dual Rectifier. My first main tackle with the 5150 is how it can be very fizzy, even when given a bias mod and biased a lot warmer (way out of crossover). I thought It could be many things along the preamp including the expressibly warm bias of the preamp tubes minus the Soldano stage.

However looking at both the 2ch recto schems and SLO 100, all three amps are nearly identical as far as the preamps are concerned (they have different ways of achieving the same and a few different values are only off by a tiny amount). The big thing I noticed was the differences in the Global Feedback, which is where the three amps really differ drastically (not that drastic but the most drastic difference). Having more feedback, the 5150 gains the ability to have more low end because with the resonance knob the signal above 400Hz (as specified in Peavey's Patent) will be attenuated much more than the low frequencies. Higher amounts of feedback also add more bandwidth making the amp more linear and "Solid State" like which would account for a lot of the fizz.

So that leads me the the MAIN question. For those that have had good amounts of time on a 5150 MkI, 2Ch. Dual Recto, and SLO 100, how much do these amps really vary in terms of characteristics? Mainly fizz and other characteristics in the high end. I know the 5150s have much more upper mid grind but have a softer high end (which some people don't like) where the other amps tend to be more cutting in the highs and softer in the mids.

The other question that I have is if anyone has excessive feedback when any three amps are cranked up. I noticed that my 5150 has excessive amount of microphonics on the first stage of V1 regardless of what tube is in that slot. I noticed that in all three schematics, the gain for the first gain stages are quite high and plan on having a much lower gain.

I will continue to ask questions as new things develop.

The concept of the amp and all of its many features where things that I wanted in an amp and deal with flaws in other amps as well of things I have heard other people mention they wished where featured in an amp. Just a bit about the amp so far:

-Built-in Germanium Diode 808-based Tube Screamer with assymetrical/symmetrical modes (footswitchable) with true bypass
-3 Channels (1 clean 2 high gain). High gain has 2 Modes each with independent tone stack and feedback controls (each tone stack can be bypassed to use an external tonestack)
-Separate High Gain and Clean preamp
-Each High Gain Mode has Independent preamp bias adjustment on the SLO stage for better lower gain non metal tones
-American and British voiced channels with each mode of both channels having its independent mid sweep and preamp bias switch
-High Gain has two gain controls like the Engl Savage and Fryette Sig:X that are accumulative in the lead mode (Gain I for the Rhythm Mode, Gain I and Gain II for the Lead Mode)
-High Gain based off SLO/5150
-Clean Channel based of 65 Twin Reverb Normal Channel
-Serial/Parallel FX loop (footswitchable)
-Second Series FX loop before the main FX loop with gain adjustment for using external tone stacks (if this is implemented the tone stack bypass will be footswitchable as well)
-Footswitchable solo boost (up to +20db)
-Stereo/Mono FX return
-External Tonestack and Global Feedback
-Power Amp line out (for recording direct)
-4 12AX7, 1 12DW7, 2 12AT7
-Any choice of Tubes, 6L6s or KT88s will come standard
-Stereo/Mono Power amp; x2 50W Stereo Power amps with footswitchable Stereo and Bridged Mono mode.
-Adjustable Selectable Voltage regulator to simulate Mesa Boogies Bold/Spongy Mains Reduction/Lower watt modes and achieve power tube saturation at lower levels
-Adjustable Primary Impedance of the output transformer coupled with variable voltage regulators to properly setup any octal power tubes from EL34s to KT88s
-Adjustable external bias with test points
-Shunted Load on the speaker outs (won't ruin the amp if a cabinet is not plugged in but is still not safe to play without a proper load)
-And best of all, ALL FOOTSWITCHING IS MIDI CONTROLLED!!! The amps will actually come with midi footswitches designed specifically for a cheap solution for each particular model.
-Full programmable Midi footswitching system, Rackmount amp switchers, FX loopers including a stereo/mono looper for the stereo return FX

Any ideas or advice is always welcome.

EDIT June 8: So I took out the idea of a preamp out, no real use for it considering the poweramp out and made adjustments to the power section's mono/stereo control, which would bring down the price a lot. I also removed the possible idea of having a reactive load attenuator as with the Variable voltage regulator, power tube saturation can be had at bedroom volumes so there is no point. There will now also be a serial FX loop that bypasses the tonestack that allows for external tone stacks at the same time there will be a bypassable global feedback output which will allow for external feedback controls as well (this would allow for an entire rackmount external tonestack and feedback controls).

I have the schematics for the preamps and power amp section. The only thing left to do is to get the tonestack and FX loop design down on the high gain part of the amp and interface it with the power section.
 
Jeebus, quite a list there. All the best in your adventure, make sure to take lots of pics along the way when you go to build it because a lot of people love to see that kind of stuff :)
 
The SLO is quite a different animal than the 5150 and Rec's, even though sharing the same basic circuit. I directly attribute this to the loop of the SLO, being pre tone stack and always on(that's 2 cathode followers that the signal is going through before hitting the tone stack and phase inverter). Add in the fact that the iron in the SLO is top notch, makes a huge difference on the final tone.

I would look at borrowing the SLO negative feedback circuit. That's always been my favorite way of doing it. Some people don't like how the presence control works on it, but I love it. Add in the depth control that Mike uses, and it sounds great.

Looks like you have a good design already. Just spend a lot of time on the layout. With all your amp will have going on, component placement will be critical.
 
you mean the DC coupled cathode follower(s) on the SLO. You threw me off by saying cathode followers because they are not straight forward cathode followers. I have so many tubes in my initial design so many that the solo boost will have to be done if i would like to keep the number of preamp tubes under 8 and I wanted to utilize a 12AU7 for the CF slots so my initial plan was to use the traditional CF for the tone stack and loop (keeps it to one AU7). Now thinking about it I do have an extra gain stage as a result of using a whole new tube for the tone stack and loop so I could utilize that unused AX7 to directly couple the stack's driver. I still do not full understand what the hell PV was doing with their tone stack driver nor do I know what the hell its doing to the sound, but I can imagine that it could dramatically effect the tone, but I have my doubts about it being anything as dramatic as the amp sets from similar circuit designs.

I will be experimenting though, I will build a Frankenstein-able prototype so that I can rewire circuits to try everything. Right off hand I know that I will be messing around with the use of Op-Amps in voltage follower (cathode follower) positions especially the FX loop to help keep the tube count low and I will be messing around with the level of feedback to find what works best. I like the idea of the SLO feedback only because it is more simple and from experience in differential amplifiers, simple is a good thing. Then again I don't think the PV is any more advanced the schematics just make the thing look a bit more intimidating and harder to read.
 
I still do not full understand what the hell PV was doing with their tone stack driver nor do I know what the hell its doing to the sound, but I can imagine that it could dramatically effect the tone, but I have my doubts about it being anything as dramatic as the amp sets from similar circuit designs.

We kinda came to the conclusion that PV was trying to sell Eddie on an active tone stack, like the one used in the Ultra series. Looks like they had it all rigged up and ready to go, then at the last minute decided against it and went with a standard tone stack. They probably already had to much work put into the amp as it was, so they just didn't modify anything else. I guess they liked how it ended up, and just kept it in the future designs.
 
I've always toyed with the idea of the Diezel 'mid cut' circuitry, which is a gain stage after the tonestack. Never had time to try it, tho', but it might be cool.
 
I've always toyed with the idea of the Diezel 'mid cut' circuitry, which is a gain stage after the tonestack. Never had time to try it, tho', but it might be cool.

I was toying with that idea too, and I actually had a mid cut/boost circuit as part of my final Hellfire schematic, but unfortunately I couldn't get the control to fit into my space requirements on the front panel and on the board. I might try it out when I design my 3 channel amp, since I'll have more room to work with.
 
The onyl change I would make ....

Clean channel should be AC30 or 100% Solid State. Not a fan of fender cleans honestly ....I know I am weird.

As cool as that would be, its not possible for numerous reasons. To try to follow accuracy would require its own preamp section, phase inverter, power section with EL84s which means another OT would be needed

I was toying with that idea too, and I actually had a mid cut/boost circuit as part of my final Hellfire schematic, but unfortunately I couldn't get the control to fit into my space requirements on the front panel and on the board. I might try it out when I design my 3 channel amp, since I'll have more room to work with.

I knew as soon as my idea started to become more advanced that I was going to have to use the 3+ channel chassis even though I consider is a two channel amp, I guess it isn't at this point.

I hate the double gain pot :D

The double gain was for the fact that instead of dropping the gain by a fixed amount like most rhythm channels do, allow the amount to gain attenuation to be user adjustable...Don't want to deal with it, turn Gain II up all the way and use Gain I to adjust both channels (this would make the gain operation the same as all other single gain amps. I am just trying to get a way for the user to get tonal colorations of different tubes. I could as well, have a Gain I for the rhythm channel and have a separate Gain I and Gain II for the lead channel. Now the question is who I skip the SLO stage on the rhtyhm channel, removing the adjustable bias on the rhythm and increase the output swing on the SLO stage (like the Savage 3rd stage). I guess that will be another experiment in the prototype to see which one works better.

Now I got the wise idea for the tone stacks on the high gain channel, giving the rhythm ch a Marshall (British) valued stack and the lead ch a Fender/Mesa (American) valued stack...or maybe vise versa and even run different Global FB to represent each flavor. Since the tone stack can be removable and augmented through external stacks, if you say want the other stack for that channel, it can easily be accommodated.
 
Here is a diagram of the front panel:

Front%20Panel.png


I have opted to duplicate the same high gain channel, but swap out tone stacks and Global FB to give more tonal colorations. Now both rhythm and lead modes have an American and British voicing Blue and White channel respectively. So not only will you be able to have British and American voicing but each voicing gets its own eq. The cool part also is the fact that for each voicing, every channel gets its independent bias switch and mid sweep for each mode.
 
Man if you're throwing the the kitchen sink, why not make the diodes switchable Ge/Si/LED to give a variety of clipping.

I know I say this every time someone does an amp design, but how about parallel V0 stage.

What are you going to run for the power section? Switchable El34/6L6 like the Mesa's?
 
Very doable with the OD part.

Parallel V0?

Power section I think will be 6L6 (or KT88) but will actually have the ability to use every octal pentode in the book.
 
If you pull this off, I'll commission you straight away

Haha sounds good.

Because this amp will be so damned expensive I think to get higher quantities of amps out and a line that is more accessible to a larger crowd. I am thinking the same circuit clean and one high gain, 100w (one American voiced with 6L6s model and one British voiced with EL34s moedl) without as many physical mode buttons (will be midi switchable only). Hopefully I will have a bottom line amp around 700-900 USD.

I just realized that I came up to par with the number of ports that are on the microprocessor, meaning I can't add anymore midi switching features without removing some features. Hopefully I don't pickup any bright ideas or some features will only be accessed via midi.
 
Sorry, V1. I always base0. Parallel the first gain stage, split the cathode/cathode bypass and bias each half differently. Should lower the noise floor and give a bit more gain. I believe the Matchless Spitfire and VHT Deliverance use a parallel triode in the first gain stage. You're too late in your design for a change like that though. I'm going to try it on my plexi build first.

Looks awesome so far though. I can't wait to see build shots.
 
Sorry, V1. I always base0. Parallel the first gain stage, split the cathode/cathode bypass and bias each half differently. Should lower the noise floor and give a bit more gain. I believe the Matchless Spitfire and VHT Deliverance use a parallel triode in the first gain stage. You're too late in your design for a change like that though. I'm going to try it on my plexi build first.

Looks awesome so far though. I can't wait to see build shots.

I planned on just giving the first stage less gain than usual and just not dropping the signal so much into the next stage, that way its not working as hard. On my 5150, the gain on that stage is so high that the 1st tube is microphonic, so i think reducing there gain there would help the issue.
 
Hey Wolfe, would you be able to pinpoint where the 5150 gets its excessive mids from? Just so that I know beforehand when I start actually building, I think it might have to do with the tone stack modifications on the Marshall stack. I could easily find out by moding the stack on my 5150 to see what it does. Also...since my knowledge of parallel RC (Shelf) circuits are minimal, what the hell is that filter network between V1a and V1b doing other than eliminating bass at lower gain settings (a fixed bright switch?). As for the Rythm filter network, I can't for the life of me make any sense out of it, the schematic is written pretty bad.

And what is the SLO depth you talk about, its not on the schematic, so I am curious as to how Mike achieved it.