Micropower subminiature tube guitar amps with reverb and vibrato

Copyright © 2005-2012 Author: Dmitry Nizhegorodov (dmitrynizh@hotmail.com). My other projects and articles

1.   Intro

The goal of these SPICE projects is to come up with one of several candidate schematics for subminiature tube-based micro-power guitar amps that are feature-reachm namely contain tremolo (a.k.a vibrato) and reverb; a tone stack and some multichannel capabiliyes are desirable too.

Three candidates emerged from what appeared to be three ill-shaped but distinct flocks of design ideas; All three are presented here in the next section; all three designs are moving targets.

2.   Designs

This section contains complete, functional amp designs alone with brief description. The next section contains additional data, measurments, simulation comparison, discussions.

All designs assume that twin-triode subminiature tubes 6021 or 6021W are used. Other tubes can be used instead, with great (and greater) success, as mentioned in the discussion section.

the designs differentiate mainly in power amp area, which in turn infuences other parts.

All designs at the moment share the same, tube-based, 3-triode spring reverb module and the same 1-triode-based tremolo (vibrato) oscillator.

All designs were tested and tuned in SPICE and in flesh (but not necessarily with subminiature tubes - check the discussions section for details) and most took many iterations.

2.1   4-tube push-pull amp with paraphase splitter, reverb and vibrato.

This design has the output stages originally influenced by popular DIY projects with self-split output push-pull triodes. Yet it has become a paraphase output stage design, which measured, SPICE-simulated, and most importantly, sounded better than self-split one.


max open gain @ 1khz in 8 ohm .................120
max undistorted power @ 1khz in 8 ohm, mW .... 650
max distorted power @ 1khz in 8 ohm, mW ..... 1250

2.2   5-tube push-pull amp with concertina splitter.

this design grew from a flock of vintage tube push-pull amp schematics all featuring bias-modulation vibrato and reverb that is *not* standard fender splitter-mixer with 4.7meg/10p gain-soaker node in the middle. Gibson GA-19RVT is the best representative. Gain is the theme of this design. Concertina splitter fits the bill, as it is ~ 2x more gain than a typical differential stage.


max open gain @ 1khz in 8 ohm .................220
max undistorted power @ 1khz in 8 ohm, mW .... 800
max distorted power @ 1khz in 8 ohm, mW ..... 11250
only 9 triodes are shown. the 10th can form an extra gain stage, see the discussions section.

2.3   5-tube push-pull Fender derivative

A 10 triode amp with classic fender tone stack, vintage fender-style sugnal mixer and typical fender power stage toplogy with presence pot. In other words, this is a derivative of classic fender vibro reverb toplogies that have a differential power driver stage, bias (not LDR) vibrato and the onmipresent reverb splitter/mixer node found an all fenders except GT16.


max open gain @ 1khz in 8 ohm .................250
              @ 100hz and 2khz.................700
max undistorted power @ 1khz in 8 ohm, mW .... 800
max distorted power @ 1khz in 8 ohm, mW ..... 1250

3.   Discussions

3.1   Selection of tubes, operation points

These schematics were optimized in SPICE with 6021/6021W in all positions; While 6021 appears to be a good "universal" triode, its versatility is a compromise: its Rp is somewhat too high for output/reverb stages and its mu is somewhat too low for gain stages. Hence an amp built with, say 6n16b in output/reverb stages and 6n17b in gain stages will function better, providing more total gain and better driving ability. Only a few parts need be tweaked to accommodate such tubes; output stage cathode resistors need be lowered by ~ 30%, cathode resistors of the gain stages can stay the same or be slightly lowered, and those plate resistors that are under 100k increased to 100k. A few minor tone adjustments may be needed for the grid shelving networks to reflect higher gain, higher Rp and higher Miller capacitance.

The following are the variants with 6n17b and 6n17b that were tested


other: updated soon...

3.2   Power dissipation

A few words on power dissipation: subminiature tubes a much less prohibitive regards power limits than normal tubes. It is not recommended to overheat or abuse them in any other way. 6021 and 6021W double-triodes add some extra confusing as different vendors and years of production provided different power dissipation limits. Roughly, all datasheets except RCA and Sylvania 1957 6021 specify 0.7W per triode as limit. RCA and Sylvania 1957 specify 1.1W per triode. Russian tubes such as 6n16b, 6n17b specify 0.9W per triode.

3.3   B+

The 230V B+ of the output stage can be combined with the 190V B+ of the reverb stage; on my prototypes these were fed from separate PS units and the two show up on the schematics. Note also that while all tubes discussed here worked flawlessly with as much as 220V plate-cathode (with 230V B+) in the output stage, and a number of proven designs (ZVex nano is oen example) also sucessfully use as much as 230V of B+, most datasheets for 6021 varaition as well as for 6n16b variants specify somewhat lower maximum permitted voltages; the actual values vary from 170 to 200. If it is a concern for you, use no more than 200V; also, any variant of these amps with a switch mode PS can greatly benefit from even somewhat lower values; providing the cathode resistors in the power stanges are adjusted (lowered) to sustain 0.8-0.85W of power dissipation per triode, no changes are needed for B+ being as low as 170V. A very satisfactory SMPS delivering 170V and 15-20ma can be built using a 555 timer or a specialized chip, a power mosfet, a coil or a transformer from a cell phone charger and just a few parts.

3.4   Plate curves and models

Any in-depth analysis of the schematics is very difficult without a look at plate curves





Lastly, composite push-pull curves:


3.5   Clipping/Cranking

The amp from section 2.1 clipping at 1k


The amp from section 2.1 clipping at 100hz


The amp vibro reverb concertina clipping



The amp vibro reverb concertina clipping at 100hz


The amp vibro reverb longtail clipping, presence dialed at 10

6021-vibro-reverb-longtail-clip-pres1.gif this one needed around 2x more imputs swing (dip at 1k at tone control

The amp vibro reverb longtail clipping at 100hz, presence dialed at 0


The ga19rvt clone amp with presence cap=20u

6021-rvt-presence-cap-20u.gif this helps to understand how the vaule of presence control pot affects tone (frequency responce). Th plot curves are for different positions of the pot - at 0, 25, 50, 75, 100 percent.

4.   Vibrato? Thremolo?

All amplitude modulation circuits shown and discussed here are tremolo. In vintage guitar amps, however, the word vibrato was used for tremolo [VIBRATO] and since these topologies derive from vintage tube amps, I follow the vintage terminology here. Note that the reverse [mis]usage was also widespread in vintage guitar world: the venerable thremolo bar prudces vibrato effect and a proper name for it is vibrato bar, see [VIBRATO] again for details.

5.   References

[VIBRATO] http://en.wikipedia.org/wiki/Vibrato_unit

Author: Dmitry Nizhegorodov (dmitrynizh@hotmail.com). My other projects and articles