Looking for single ended build ideas? Maybe w/ KT88??

[Jalmeida]

Hey,

So as an exercise to learn amp math in design I want come up with a single ended amp with 2 tubes, maybe 3 max(with rectifier tube) and was looking for ideas. I have been researching and looking at designs to familiarize. Found plenty of ultra linear Hi-Fi KT88 designs, but not much for single ended guitar with KT88.

I just want to research designs, play with some variations and learn all the math parts of the design. This is part build and part learning exercise. Idea being I can prototype, build and tweak also and if it is crap it never becomes anything more than a learning exercise. No kit builds. Thanks. 

[kagliostro]

Just a suggestion as to save a bit of money, don't use a tube rectifier, SE amps don't has benefit from them

An SE amp don't SAG if you use a tube instead of a Solid State rectifier because there is a negligible current change from idle to full power

Franco

[PRR]

A easy-working SE 6V6 amp is LOTs cheaper than a full-blast KT88, and not THAT much less loud on the ear.

I labored on a SE 6550 aiming for 17 Watts. A 3W-4W 6V6 would be much cheaper, lighter, more common parts, and better than half the volume.

The math-work is about the same either way. Even for a third-Watt 6AU6 amp. (Actually the 6AU6 does not publish good power-amp conditions so I had to do more thinking and plotting.)

[kagliostro]

A quick search pointed to this


Franco

[kagliostro]

If you foollow PRR council (his advice is always very wise) you can find lot of projects ready to be realized

Franco

[shooter]

Found plenty of ultra linear Hi-Fi KT88 designs, but not much for single ended guitar with KT88.

I built a few KT88's, I started ^^^^^ there for the PA, then added preamps stole from known good guitar amps, glued them together, walla.
I've found in SE land you want a PA section that will Amplify, without a lot of "Tonal" flavor added, get that from pedals, preamp-design.  Keep the PA section simple.
some ranges that I like;
The 88 biased real close to Max plate dissipation
B+ between 400-450 at plate
an efficient speaker that can handle Min 50w rms
personal choice - UL OT


and as PRR said, disposable cash

[bmccowan]

If you want to take PRR's advice, and I would even if he made me do 200 Ohm's Law math problems before breakfast, but don't want to build a standard Champ (because everybody does that, or some other reason.) You could look at some similar amps:
The version of the Gibson GA-5 that is almost identical to the 12ax7>6V6 Champ.
Or several Valco SE 6V6 amps - the 6SL7>6V6 being a favorite of mine and others.
I have not built a SE KT-88, but I did build a SE 6L6. It did not become cool/different until I added a Framus type mid control which gave it its own flavor.

[Jalmeida]

A easy-working SE 6V6 amp is LOTs cheaper than a full-blast KT88, and not THAT much less loud on the ear.

I labored on a SE 6550 aiming for 17 Watts. A 3W-4W 6V6 would be much cheaper, lighter, more common parts, and better than half the volume.

The math-work is about the same either way. Even for a third-Watt 6AU6 amp. (Actually the 6AU6 does not publish good power-amp conditions so I had to do more thinking and plotting.)

My aim outside of learning if all goes ideally is to also have a basic amp in the >12w range. Then from there long term goal is have a circuit that I can play with by adding switchable channels and cascade gain stages etc. I have found lots of single ended EL34 and 6L6 amp ideas. The KT88 kagliostro shared was the first non hifi ultra linear type I have seen.

[pullshocks]

Dummyload and Shooter have posted some whomping SE designs you might be interested in. 


low watt 6L6GC practice amp. (el34world.com)


Since It Seems To Be Champ Amp Month... (el34world.com) 

[sluckey]

Here's another se idea. Quite different from a champ. All pentodes. Probably couldn't fit a KT88 though.   

     http://sluckeyamps.com/smoky/smoky.htm

[Jalmeida]

Here's another se idea. Quite different from a champ. All pentodes. Probably couldn't fit a KT88 though.   

I love the toaster idea 😂
     http://sluckeyamps.com/smoky/smoky.htm

[pullshocks]

And then there is the THD Univalve

[Jalmeida]

Found plenty of ultra linear Hi-Fi KT88 designs, but not much for single ended guitar with KT88.

I built a few KT88's, I started ^^^^^ there for the PA, then added preamps stole from known good guitar amps, glued them together, walla.
I've found in SE land you want a PA section that will Amplify, without a lot of "Tonal" flavor added, get that from pedals, preamp-design.  Keep the PA section simple.
some ranges that I like;
The 88 biased real close to Max plate dissipation
B+ between 400-450 at plate
an efficient speaker that can handle Min 50w rms
personal choice - UL OT


and as PRR said, disposable cash

What kind of Wattage were you able to tweak out of the KT88 SE builds? I have been working on drawing the loadlines at 400v, but not sure yet where I will be for wattage.

[shooter]

Probably couldn't fit a KT88 though

see reply #8, my take on your juke-box

What kind of Wattage were you able to tweak out of the KT88 SE builds?

16 - 23W RMS at the speaker depending on how hot bias/abusive you are
played through a 2X12 with low 90's SPL speakers I couldn't convince a single guitarist though, they ALL said it was 30W, who am I to argue with them
 

[tubeswell]

What shooter said. Min 400V on plate. Bias for Pmax. Keep the screen about 50V lower than the plate. Add your own preamp.

[AmberB]

As for the KT-88, there's tube data for the conditions for 20 watts for a single ended tube.  I was thinking about building one but I didn't want to spend that much money on the output transformer.
I've built 2 different SE amps with a 6L6GC using the transformers from 10 watt PA amps.  They work well and get pretty loud through a 12 inch speaker.

[allstonma]

Id suggest an ab764 VidroChamp with a KT88. Built a couple for people and they seem to like them.
Just my $.02....
TH

[labb]

Don’t know about the learning experience but the SEL from the old AX 84 site is one heck of an amp. If you look deep enough you will find that the original design was to use either 6v6, 6l6 or kt88. The schematic ended up just showing the KT 88. I built that amp with a way to externally adjust the cathode resistor. Also with socket for EL84. Fun amp. After all of the playing around settled on the 6V6. I’m thinking about marrying the preamp section to a 20 watt PP power section just for the fun of it.

[Ended-single]

the SEL from the old AX 84 site is one heck of an amp.

I know I'm late with this comment, but I wanted to second this.  The SEL is a fantastic place to learn and the resulting amp I made is my #1 and I think it sounds fantastic.   You can tune this circuit yourself and make the amp have the frequency response and gain structure that you want. It may seem like a stale design because it doesn't seem to get talked about much these days but I am telling you that it really is one heck of an amp.

[Jonas]

More info here.....

https://el34world.com/Forum/index.php?topic=31067.0

[Willabe]

Like PRR and others have said, the thing is with higher output SE amps is the cost compared to PP amps of ~ the same power.

Not only is a higher wattage OT much more expensive to get to 15w to 20w, but there is a considerable size and weight penalty.      But a pair of 6V6's and their OT can get you ~ 20w/22w for a lot less $$, weight and size.

And if you build a KT88 ES amp it's not going to sound very different than a PP ~20w amp if your playing on the clean side. To really hear the sound of that KT88 you have to push it into distortion. Will you be able to turn it up that much? 

Not trying to squash you vision. It is a lot of fun to build an amp and you always learn things from building it, just want you to think it through.   

I built a Princeton 5F2A, which is a Champ with a single knob tone control, think tweed Deluxe tone control. I wanted a lower output amp for at home and I wanted to/feel what a SE amp sounded like, and see if I like them. And like PRR and others have said, it's still plenty loud, sometimes still too loud.

I built it with a rectifier tube so I could change the dcv by swapping rectifier tubes, 5Y3, 5V4, GZ34. I wanted to hear the difference at different dcv's. I built it with a 10w SE OT that is made by Haybor, size/weight/$$ were just not that much more than a standard Champ type SE OT. I was running it through different single 12" speakers I have and through a single WGS alnico 15". (Always run through a Fender stand alone reverb clone I built.)

As you would expect, the 6V6/5Y3 breaks up the fastest and the 6L6GB-C/GZ34 is the cleanest. But the 6V6 sounds pretty different than the 6L6, wither both are clean or both are ~ the same break up, 6V6 is warmer/fatter/less sparkle, tried with both NOS and new power tubes.

Problem is the extra heat when using the 6L6/GZ34 combo, that steel tweed chassis gets HOT! in the tweed Princeton 12" speaker combo cab. Could be fixed with a small fan in the cab, even a small pedestal fan set behind the amp.

If I built a SE amp again, I'd go with SS rectifier and put in Power Scaling/VVR  to adjust the output tubes plate/screen DCV. I can see a 2nd Power Scaling/VVR just for the preamp set up  with a dc voltage clamp on the preamp tubes dcv, so as you dial down the power tubes dcv and unload the PT's B+ dcv, it clamps the peramp tubes dcv at where it's set, it doesn't let the pre dcv rise. And I'd want to build it to use 6L6 types including KT66 and EL34/KT77. Just to hear them.

I'd use a full AB763 preamp, with full TS including mid control, so 2 preamp tubes, probably  have the input 12AX7 wired up to be either single or in ||. You would have several different options this way, like a 12DW7 in ||, which would be 1-12AX7/1-12AU7, or a 12AU7 in ||, or a 12AY7 in ||. A plus of the full tone stack would be to help eat up some gain from the 3 stage pre. A volume and and a master volume.

And the amp build I've been thinking about for several years would be a stereo SE 6L6 with the Magnatone stereo pitch shift vibrato. Get around 16w to maybe 20w? I have 1 of the new stereo Maggi's and it sounds wonderful to me, but it's 42w's with a 2x12"s stereo cab. A little much for at home

Magnatone does make a lower wattage stereo amp but it's EL84's and the 2x10" speaker cab has the speakers to close to together to hear the watery slap of the sound flipping speakers back and forth. You'd have to hear to know if you like it. I love it! Never heard anything that sounds like it, and a Leslie is a different animal, wonderful sounding, but a different animal than a Maggi.   

[Ended-single]

More info here.....

Very useful information.  I went through the champ build, a bassman, and one more before I finally put the effort into a SEL build.  The early ones helped get my skills up for the SEL build.  I used a modified circuit (I imagine everyone modifies it in some way, that's part of the charm).  Testing different components and variations and understanding how to control the tone, gain and frequency response is the key to making it 'your sound'. 

I'm going to throw something at y'all here and let you know that my approach to single ended amps is based on running into an active load box with transformer isolated output and not into a speaker at all.  I have been using a variation of Aiken's active load schematic that I'll share as an attachment.  It's similar to the Ox box thingy.  It makes the transformer think it's driving something close to a real speaker.  It's a night and day difference form driving a resistor as a load.  The use of speaker crossover components creates a resonant peak in the impedance curve plus an impedance rise at the treble end.  You can measure the resulting impedance curve of the active load using the REW software pretty easily. 

In this circuit there is a big 'ol wirewound resistor and two other sections, one as the "treble inductor" which makes a rise in impedance in upper frequencies and the other (the more important section) is the big impedance peak created with a cap and a coil.  Both of these sections have bypass resistors that are not required but are useful to tune the magnitude of each section.    The schematic I'm showing does not have the correct values and isn't exactly what I used, but it's very close and gives you the idea - I can't locate my actual schematic at the moment but if there is significant interest I'll try to give more detail.  My purpose here is to try and get some of this info out there because it's been working so well for me and sounds so good compared to the other resistive attenuator widgets out there.

also attached in an image you will see how I build a test kit that allows me to tune the circuit to any tube amp.  I put in multiple coils, multiple caps, and variable resistors for all of the key components.  In this way I tune by playing the amp and then make the circuit that is optimized for that amp.  In general I find that a 13-ish mH coil and a 120 to 160 uF 250V cap.  I use a 10 to 16 ohm wirewound resistor.  In the schematic it shows 20mH/5A inductor and 200uF/400V cap, but that's overkill for an SEL so you can bring those down to something more reasonable, but don't cheap out here. 

-continued in next post -

I have attached a pic of my load box and a rough schematic, I'll make sure they show up properly.

[Ended-single]

The quality of the wirewound resistor, the coil and the big capacitor is critical.  You have to use decent quality components even for a small SEL amp.  Do not use electrolytic caps.  Do not use cheap tiny resistors, and get a proper inductor.  I used to use all erse products, but they shut down.  Good inductors are here (where the US erse folks moved:  https://uscoils.us/  But they don't carry the really big caps anymore.  Parts Express options are quite expensive but there is probably other options for affordable ones.  You want a 250 V or higher polypropylene or metallized poly cap.  They are like $50 to $120 each at parts express, so yea, you have to put some money into it.  I tried all the other cap options, they all caused the touch sensitivity of the amp to go away.  Same with the big inductor.  I used 8 to 13 mH typically, like this:  https://uscoils.us/product/ussq55-14-13000/   Also costs a bit of money.  I do make these to handle high wattage so I think one could take a step down in cap voltage and inductor cost, but my point remains that the component quality matters here.  For the big resistor I use proper Ohmite ones, 50 to 200 watts depending on the build.  I like to get the ones that are variable to allow for tuning the circuit, but the differences get subtle so a 10 ohm fixed vintage ohmite from Ebay is a simple way to go.

Get a decent 10k 1:1 transformer for the isolated output.   I then go into a Redbox but there are lots of options like IR boxes and such that you can use to tap the signal at the load box and convert it into a balanced (or unbalanced) low impedance output with the proper bandwidth limiting applied so it sounds like a guitar speaker.    The loadbox is great for the SEL because you can control the "master volume" AFTER the amp.  The idea is that you run the amp wide open on the back end and vary the input pot.  Then you take the output of the amp (running hot to get that power tube going) and run it into the load box.  The output of the loadbox is controlled and you get exactly the volume/level you want while allowing the amp itself to run at the power needed to get the tone you want.

The SEL is very quiet and I can record direct with this setup and it sounds great to me.  YMMV vary of course.  I just thought I'd share because this info wasn't out there when I was trying to do this build and what I've included here are the critical factors needed to build a nice quality all analog reactive load box with transformer isolated line output.

[Willabe]

Here my Princeton 5F2A, faux Gator. 

I sent Mojo the tolex and grill cloth, they built it and covered it.

[mresistor]

Here my Princeton 5F2A, faux Gator. 

I sent Mojo the tolex and grill cloth, they built it and covered it.

Man that looks fantastic!   Beautiful..        Congrats

[tdvt]

my approach to single ended amps is based on running into an active load box with transformer isolated output and not into a speaker at all. 

What is your signal chain after that?

[Ended-single]

What is your signal chain after that?
[/quote]

After the active load box you basically have an unbalanced signal that can go back into guitar pedals or other line input devices.  The transformer isolated output of the active load box has all of the highs and lows that a typical guitar speaker filters out. I put this into a Redbox analog speaker simulator and this gives me a balanced low impedance mic level signal out that has the filtered bandwidth of a typical guitar speaker.  Alternatively you could go into a IR pedal instead of the Redbox and get a similar result.  I also have a send/return at line level just before the Redbox that I can use for reverb and a parametric EQ pedal.    I most often send the signal from the active load box output through a Broughton parametric EQ (reverb goes here if you want it) then into the Redbox and the XLR out to a mixer or audio interface.  I can then monitor through a stage monitor, studio monitors or IEMs.

To put it another way, the output I get is not unlike the output from a typical amp sim pedal.  Guitar into the amp (not simulated), then line level raw amp signal out.  On the amp sim pedal you choose how you want to do the speaker sim and I do the same, just with an analog Redbox.  A lot of people are working with guitar direct through these pedals and this is a similar approach but with a real amp instead of the simulation.   

[Ended-single]

Attached is an example application of this approach.  The THD univalve has one of these transformer isolated line outputs built into it.  What I am showing here is the simulated speaker that I plug into the amp output in order to provide the active load.  What you see here is the essential components in an active load plus the Redbox speaker sim (which you can replace as you see fit with your own IR speaker sim).  The power supply on the left is to supply power to the Redbox.  The key components are the inductor (I believe that's a 18mH), a 160 uF 250V Erse capacitor, a smaller coil for the HF  and  two small ceramic wirewound resistors that act as the bypass controls to scale the impedance curve appropriately.  The big resistor is a vintage Ohmite at like 100 or 200 watt rating with a tap at the appropriate ohm level.   All values were determined with that device I showed in a previous post.  I build this into a box that sits under the THD with an XLR out as shown in the second attached photo.  Guitar in, XLR low impedance mic level out.  No speaker.  On the front of the box (not shown) is a 'master volume' that controls the line level signal post isolation transformer and jacks for send/return after the master volume. 

In terms of functionality, it's a rather bulky full analog amp non-simulation pedal really.  The digital ones are much smaller and lighter, but I like the all tube signal path.  You would be surprised at the great dynamics you can get out of this rig into a wedge monitor.  The SEL is a big step up from the Univalve top me because I was able to tune the circuit to give me the response I wanted.  Keep in mind that this active load approach can be applied to _any_ tube amp.  Those components would easily handle a higher powered amp.  I use lower wattage amps because I like the single ended sound and I just don't need the extra power of a higher wattage amp since I'm not actually powering a speaker. 

One friend used this approach to capture his amp for a Tonex (or some other modeling device, maybe it was a Kemper).  He said the difference in amp capture using this load vs a purely resistive load was significantly different (as in better).  It's not a real speaker, but it's a heck of a lot closer to what the amp wants to see than just a big resistor because it's simulating the variable impedance vs frequency that exists in a regular speaker.

[Jonas]

What shooter said. Min 400V on plate. Bias for Pmax. Keep the screen about 50V lower than the plate. Add your own preamp.

With a single ended KT88 - is keeping screen voltage about 50V lower a safety design? What was the reason for this again?

[tubeswell]

What shooter said. Min 400V on plate. Bias for Pmax. Keep the screen about 50V lower than the plate. Add your own preamp.

With a single ended KT88 - is keeping screen voltage about 50V lower a safety design? What was the reason for this again?

If you’re aiming for centre bias idling at Pmax with a load resistance of Zout=Va/(Pa/Va), then you’ll likely end up with a load line that crosses below the knee of the Vg0 curve, in which case, keeping the screen about 50v below the plate will lower the grid curves enough to prevent too much screen current when you dime the amp with power chords (of course, there are other things you can do instead, but that’s just my preference. YMMV)