Category Archives: D.I.Y. Amps

sof86_t_01

Spirit of 86: D.I.Y. Tube Guitar Amp Head

Comments are disabled because I got tired of the endless spam. For questions, comments, and/or dilemmas, email th@thallenbeck.com.

Here are two videos of the Spirit of 86. The first one is done with a Fender Telecaster and the other is with a Gibson SG. Scroll down for photos and an erudite philosophical discussion.

sof86_front_02sof86_chassis_03

Spirit of 86 schematic

This is the Spirit of 86… or as I like to call it, the Buttrock Express. It’s like a little shrine built around an EF86 tube (http://en.wikipedia.org/wiki/EF86). A couple of years ago, I started wondering why I knew of a number of amps that use an EF86 as a first preamp stage, but I’d never heard of any amps that use a preceding tube to drive an EF86 (of course, that doesn’t mean there aren’t any). I thought it might have been because it would just sound horrible or because resonance and microphonics would get the best of the EF86, but I really wanted to find out. So after dawdling around for several months, I put one together last year, with parts and components I happened to have handy, and I tweaked it every few weeks until a few days ago when I decided to immortalize it in a blog post and stop fiddling around with it.

Spirit of 86 layoutTo the right are a schematic (above) and a diagram for a possible layout which is close to the one I used for the amp. Of course it’s not the only possible layout, and the positioning of the elevated heater supply (upper right) is arbitrary – I wound up kludging it onto the end of the turret board.

In the layout, the ground connections (thick blue lines) are an approximation, but most of the points that connect to a specific location such as the from the Gain control to the turret board are drawn that way for a reason – to reduce noise – because this project was probably the most noise-prone one project I’ve done to date. The biggest problem was AC hum. Randall Aiken has a good article about grounding in tube amps and why it’s not always good to perceive the ground as a sort of uniform field:

http://www.aikenamps.com/index.php/grounding

… and I try to do whatever the voices in Randall Aiken’s head tell me to because it almost always pays off. Side note: I tried grounding the EF86 tube to a couple of different points on the turret board instead of running the ground back to the star. I really thought it would make a difference in noise level but didn’t, to my ears anyway.

sof86_interior_02sof86_interior_01

This project has been interesting and educational for me, but I still can’t decide if I like the way it sounds or not. The overdrive is aggressive and “gnarly” for lack of a better term (which makes sense because the EF86 is a pentode), but it doesn’t have a lot of sustain or the “squishiness” of cascaded triode stages. I suppose that might be a good thing for articulation and clarity if this amp were being played by someone with better guitar skills than I possess.

The Buttrock Express might serve as a template for an overdrive channel at a future date. Getting a nice loud clear clean tone out of it is an exercise in futility.

The Thomas the Tank Engine sticker on the front face is there to cover a drill hole that I would up not using. The TH Audio badge is there because I had some extras – this is just an experiment, not something I’ll be offering through TH Audio (my side business for effect pedals and, eventually, amplifiers).

sof86_front_01sof86_chassis_02sof86_back_01sof86_chassis_01

Controls are, from left: Bass, Mid, Treble, Gain, Bright, and Master Volume. The Bass/Mid/Treble controls are a fairly standard Marshall-style 3-band EQ positioned after a cathode follower. The EQ stage uses a couple of unusual values: 68k for R7 (see schematic) and 250k log for the Bass pot (instead of 500k or 1M, to counteract the mounds of low frequencies that pile up). The Bright and Master controls are connected near the input to the power tubes, like one might see in a Vox AC15.

The amp dishes about 20 Watts, with a 6.6 kOhm output transformer and a 6L6 push-pull pair. At 20 Watts, the 6L6’s aren’t doing much, but I keep them there because a) the OT I’m using is a nice match for them (better than the 4k OT I was using previously, b) a 6V6 pair sounded glassy and midrange-y, and c) I wanted to hear the EF86 overdrive without an extra layer of power-tube breakup.

I used a tube rectifier instead of some diodes because… well, because I could, I guess… although the tube rectifier might be introducing a little ‘sag’ that mellows the general gnarliness of the pentode overdrive. To be honest, I don’t know if it does or not.

For this project, I used Tung-Sol tubes because… well, because I already had some. I’ve swapped out the Tung-Sol set with an Electro-Harmonix set, and I don’t hear much difference at all. In either case, it wasn’t difficult to get something close to 90V at the plate of the EF86, once I figured out the voltage drops. Normally, I like JJ tubes and have used them in a number of projects, but to my ears, the JJ EF86 sounds weak in this particular project (I tried two different ones).

One last thing: this amp is *not* a clone of the Dr. Z Route 66. The Route 66 uses a single EF86 into a passive EQ section, to a phase inverter and a fixed-biased KT66 push-pull pair. The Spirit of 86 uses 12AX7 -> passive EQ section -> EF86, into a phase inverter and a cathode-biased 6L6 pair (which is hardly doing any work).

-T. Hallenbeck, Oct. 2014

th@thallenbeck.com

 

D.I.Y. Ampeg B15n Fliptop Clone (2013)

Comments are disabled because I got tired of the endless spam. For questions, comments, and/or dilemmas, email th@thallenbeck.com.

Originally uploaded September 2013

This is more or less a clone of the head section of the Ampeg B15n, known as the Portaflex or “fliptop.” I’ve been wanting to try building a tube bass amp for a while and I’ve been curious to find out why fliptops sound the way they do.

B15n Clone - back view, complete   B15n Clone - schematic

B15n Clone - partial layout

 

 

 

 

 

 

Below is an exciting video documenting the tail end of this build. Please keep in mind that I’m not exactly Tony Levin.

B15n Clone - interior #1   B15n Clone - action shot

My fliptop interpretation preserves most of the “canonical” aspects of the original(s), like the extremely high input impedance (5.6M Ohms), the Baxandall tonestack, -50V fixed-bias (*not* cathode-bias!) and the unusual phase-inverter configuration. But this one sports a few departures from the original:

1. The instrument input uses a decoupling capacitor (22nF) before the biasing resistor of the first preamp grid input.

B15n Clone - interior #22. The resistor and capacitor values for the DC filtering stages aren’t anywhere close to the original(s) – I used 50uF can caps because I had some available. The difference in the filtering stages probably changes the whole sound but I don’t have a way to find out because I don’t have access to a real fliptop. I’ve seen some B15n schematics specifying diodes for rectification and others specifying tube rectification. Mine uses diodes. I wound up with about 465V at the plates of the power tubes, which might be a little too high, but I haven’t done a round of tweaking lately.

B15n Clone - top, no cabinet #13. My version has one channel instead of two like most others do, because I wanted to keep the interior as uncluttered as possible for the inevitable 5 million little adjustments I made.

4. I used a great big filter choke rated for 500 Volts, because I could. I found that a choke helped reduce noise and hum.

5. I always try to use elevated heater supplies (see schematic), even for push-pull amps. In this case, the filament divider resistors are tied to the -50V bias (see schematic). That’s one way to deal with 60Hz/120Hz hum from 6SL7/6SN7 tubes. I guess this doesn’t qualify as a “departure from the original” but it’s how I got rid of most of the AC hum, along with isolating the ground lines for the filter capacitors from the other grounds and busses, like I should have in the first place.

B15n Clone - top, no cabinet #26. I found that I could coax more clean headroom out of the unit by using 6SN7 tubes for the preamp stages instead of 6SL7 (think: 12AU7 vs. 12AX7, sort of). For bass, I like clean headroom. I looked at the signal from input to output with an oscilloscope and found that clipping was most likely to happen first in the preamp stage. That can sound nice, but for this unit, it sounded awkward and nasty, to me anyway. Swapping 6SL7′s (higher gain) for 6SN7′s (lower gain) reduced the strength of the preamp signal but gave more headroom before the nastiness kicked in. The drawback for 6SN7′s is that they demand twice as much filament current as 6SL’s (~=600mA instead of ~=300mA) but the power transformer is quite capable of handling that increase (see schematic).

My version puts out somewhere around 30 Watts at the most, from what I can tell. I used a Classictone 40-Watt output transformer, so I don’t want to try to push the output higher than it already is. To be honest, I’m guessing at the Wattage, assuming about 20% or 25% power loss for the push-pull output transformer.

The tubes are all Tung-Sol. I tried a pair of Tung-Sol 7581′s for the power tubes, and a pair of JJ 6L6GC’s, but I liked the Tung-Sol 6L6GC-STR’s for their relatively wide frequency response.

B15n Clone - interior, no board #1

B15n Clone - interior, no board #2

B15n Clone - front, completed

B15n Clone - top, no cabinet #2

B15n Clone - back with cabinet

D.I.Y. 5f2 Champ/Princeton Clone (2013)

Comments are disabled because I got tired of the endless spam. For questions, comments, and/or dilemmas, email th@thallenbeck.com.

Originally uploaded April 2013

5f2 Clone

5f2 Clone - action shot

This is my first single-ended amp build. It puts out about 5 Watts, which is plenty loud, but not bonecrushing. It’s based on the Fender 5f2 “Princeton” circuit, which is quite similar to the 5f1 “Champ,” the main difference being that the 5f2 has a tone control. This is not a kit – the chassis, cage, and filter choke are from Hammond and I kind of winged it with the drilling. The transformers are from Classictone and the tubes are JJ.

Princeton/Champ DIY builds have been done many times, but now that I’ve built my own, I can see, and hear why so many people like having at least one of these around.

If you see anything I’ve said here that might be inaccurate or wrong, please let me know so I can correct it.

5f2 schematic

Below are the schematic diagram for this build (left) and a possible build layout (right), and a video to prove that my latest squid launcher actually works. The layout diagram is close to what I actually built but not exact, and it’s not to scale – it just shows where things can go for minimal wire-crossing and halfway-decent ground distribution.

5f2 Clone schematic   5f2 Clone - layout

Take the layout diagram with a grain of salt – the resistors that connect directly to the grid inputs of the tubes (R2, R5, and R10) really should be soldered directly to the socket pins, with as little exposed lead as possible. Also, the cathode resistor for the power tube (R11) probably should be rated > 3W.


Since I’ve never gotten my grubby hands on an original 5f2 or 5e2, everything I’ve learned has come from schematics, sound samples, video clips, other people’s blog posts, and discussion boards. But from what I do know, the main differences between this build and a ‘canonical’ 5f2/5e2 are:

1. Different location for the ‘tone’ control. I put ‘tone’ in quotes because it’s just a low-pass filter. For this build, ‘tone’ is situated after the second triode just before the power stage, as a 10n capacitor to ground and a 250k pot.

2. Larger values for the B+ filter capacitors. The 5f2 used ~= 8uF whereas this build uses a 10uF after the rectifier, and 47uF for B+2 and B+3.

3. 6L6 power tube instead of a 6V6. I like 6L6 in general.

4. The 5f2 had a feedback resistor between the secondary side of the output transformer and the cathode of the second triode stage. This build doesn’t. Originally I had one but I thought it made single-coil pickups sound too harsh so I got rid of it.

5. This build has a (relatively) large bypass capacitor in parallel with the plate resistor of the first triode stage. It rolls off high frequencies and helps to alleviate the ‘icepick’ effect at high volumes.

6. This build has a 220k grid-stop resistor at the input to the second triode stage of the preamp. Like the cap mentioned in #5, it helps to roll off high frequencies.

7. This build uses an elevated heater supply (see schematic) with a DC offset of about 45 Volts for the filaments, to reduce AC hum.

8. This build has separate outputs for 16, 8, and 4 Ohms. The 5f2 usually had a single output.

5F2 Clone - complete   5f2 Clone - back view

I’m sure there are other differences I’ve forgotten to list. There is a little AC hum but it’s quickly overpowered as the volume knob goes up. Overall, the build is a little messy – I didn’t really know how it would go together when I started it, and I made a zillion little tweaks to it. I left the transformer leads a little bit long in case I reuse the transformers for other projects. Next time, I’ll use a chassis that’s higher than 2 inches, and a little wider for better component spacing – the choke just barely fits. Near as I can tell, I’m getting about 5 Watts at the output (assuming 50% loss for a single-eneded output transformer). That’s plenty of cowbell for just sitting around playing.

Like the original Champ, this build does not use a master volume control. That means the power stage is always running near full boil. As the Gain knob goes clockwise, the volume increases, but so does the clipping.

The Classictone output transformer I used (40-18031) has two different leads for 5k and 8k primary impedances. I’m using 5k here for a 6L6. 5k would also work for an EL84. 8k would be good for a 6V6.

5f2 Clone - top view

Above, from left: 12AX7 preamp tube, 6L6 power tube, 5Y3 rectifier. All the tubes are JJ. I’ve been  getting good results with JJ tubes lately, especially the JJ 6L6. The ‘sag’ from the 5Y3 rectifier is really obvious in the sound of the amp: to me, it’s like the signal is hitting a rubber wall when I lay into the strings. The knobs, from left, are volume and tone. There is no ‘master’ level control. The volume pot controls the strength of the signal from the plate of the first tridoe stage in the preamp to the grid of the second triode stage.

5f2 Clone - another top view

Below are photos of the interior. It’s pretty obvious that I had some issues with the layout.

For this build, I tried ‘floating’ the filament wiring (green twisted wires) instead of having them hug the chassis.

5f2 clone - interior #1

5f2 Clone - interior #2

5f2 Clone - interior #3

5f2 Clone - interior #4

Tube Guitar Amp #1 (2012)

Comments are disabled because I got tired of the endless spam. For questions, comments, and/or dilemmas, email th@thallenbeck.com.

Videos are at the bottom of this page.

As of April 18 2012, my first D.I.Y. tube guitar amp appears to be functioning normally. I made a huge mess and a lot of mistakes with it but it came out fairly well for a first try, and was definitely worth the time and expense because I learned quite a bit.

hta1_roughHere is a schematic that’s more or less like what I made, which is based on the AX84 designs (see ax84.com, a website for tube-amp enthusiasts). This schematic is incomplete and has some mistakes in it – the tonestack is not represented and the filament supply of the Hammond 270FX power transformer should marked 5 Amps, not 3 Amps. The transformer, a Hammond 1650F in this case, is matched better to a pair of 6L6’s than to EL34’s, and the poweramp section isn’t right for a pair of 6V6’s, which start red-plating right away.

See below for why the tonestack is left out. If you want to see tonestack schematics you can probably bring up 500 of them in Google in about 20 seconds.

ta1_interior3   ta1_interior2

The vertically-mounted perfboard is for the the power tubes’ negative bias voltage, which I wound up not using because I opted to cathode-bias them instead.

One of the dumbest things I did was to cram most of the DC filter capacitors onto the turret board. I should have used dual can caps for all the filter-cap stages instead of just the first two, so all the ground lines could be run straight back to the star ground.

ta1_underside   ta1_interior1

Since the poweramp is cathode-biased, I don’t have to adjust a negative supply voltage every time I change tubes. I’ve been able to load EL34’s, 6L6’s, and 6V6’s, all with very distinctive results, although I can’t keep using 6V6’s because they overheat in this circuit.

The EQ is not a standard 3-band stack. The bass control is just a high-pass filter, the mid is a twin-T notch, the treble is a low-pass filter. And they’re wired in series, respectively. I’m not thrilled with the EQ but a standard 3-band Fender-style EQ sounded awful with the overdrive stage, probably because of impedance mismatches and certain sections I didn’t wire up all that well.

The tubes in the photos are 6V6’s but I stopped using them because they red-plate (overheat). I used EL34’s for the videos but I’ve been using a 6L6 pair lately because they give me a better low-end response than the EL34’s. That isn’t surprising since the impedance of the output transormer (about 7.7kOhms) is way out of whack with an EL34 pair would expect (3-4kOhms would be better). I’ve got the amp wired for a single 16-Ohm output because at this point all I’ve got is a single 16-Ohm speaker.

ta1_top1   ta1_top2

ta1_back1   ta1_back2

Crunchity-crunch, with a Gibson SG:

Medium/high gain, with a Gibson SG:

Sort of clean, with a Fender Strat:

Medium/high gain, with a Fender Strat: