RGB LEDs for power indicators (again)

Originally, I had been planning to provide a "standby" switch, i.e., a switch on the high-voltage DC power, separate from the heater power (for the tubes).  There would have been separate LEDs for "power" and "HV", and at one point I thought it would be clever to have the two LEDs be colours within an RGB LED.  However, how to power the LEDs was an issue, since 5v DC power is only available when the preamp is on...

In any case, the Internet talked me out of standby switches; instead, there will just be a "mute" switch.  So it was back to one plain power LED for each of preamp and power amp, presumably red.  I planned to power these LEDs from the heater power lines: which are not referenced to ground except through the "hum bal" pots.  So, the LEDs would be fed from primitive DC power supplies (diode, capacitor) running across the two heater lines.

But especially considering my use of tube rectifiers, it's nice to have a light that comes on when the HV is active, after the warmup period.  And if that is the blue, and heater power is the red, within an RGB LED, then the colour will start out red, then shift to purple as the amp warms up.  At power off, the blue will stay lit and gradually fade as the HV discharges.  This is a lot of good information (wouldn't have to be RGB of course, that part is just a cute gimmick; could just be separate LEDs).  But to use RGB LEDs, everything must be ground-referenced (common cathode assumed).  So the circuit from the heaters becomes two diodes, one from each heater line, to a capacitor which goes to ground; the red LED resistor feeds from the junction of diodes and cap.  The red LED brightness will vary as the hum bal is adjusted.

Vibro-Champ

(Please note, these mods are as-yet untested, just a "try this" suggestion.)

a digression: Fender Vibro-Champ

This doesn't directly concern my amp design; but I have based quite a bit of the amp design on the Fender amp circuit.  Anyway, here's a related project, taking a Fender Vibro-Champ amp, and re-purposing the 12AX7 tube from the tremolo circuit, to become a second gain stage.  This frees up two pots, which are re-purposed as the missing "mid" knob for the tone-stack, and the gain control for the second stage (the original volume knob now behaves as a master-volume).

One could use a pull-switch pot for the channel switching, but in my case, I want to re-use the same components as much as possible (i.e., the original pots), and I already have a non-original switch on the back for disabling the negative feedback; I can re-purpose this switch as the channel switch ("2nd stage").  Or, one could drill a hole in the front panel; presumably this would be the least desirable option for most people.

*** Edit: scratch that above, I'm now thinking the best plan may be to put a switch in place of the seldom-useful second input jack.
(Notice that in my own input section, which in many ways copies Fender, I have a second input jack, but the resistor ratios are different, to make a bigger difference between the two jacks.  The Fender is only a divide-by-2, i.e., -3dB.)

(If it's not obvious, these modifications turn the Champ preamp into something very close to the "F+" gain stage, in my amp design.  This project is both a way to make my Champ more versatile, and a way to prototype the "F+" circuit before I actually begin construction of the amp.)

input section (update)

The main update to this section is in the cathode circuits connected to the "voice" switch: I now have the 2.7k resistor permanently in-circuit, and the switch just combines more resistance with that, in parallel.  This is so that the cathode is never open-circuited (allowed to float) as the switch is turned.

power amps and power supplies

With the following schematics, the initial design is complete.  I.e., if you collect all the schematics posted so far (PNG images), and eliminate older versions which were superseded, then you'll have the schematic for a full guitar amp.  Of course, some values are unspecified ("TBD"), and no doubt there will be many changes and additions to come.

power amp design -- Lafayette LA-224

The previous post basically concludes the (initial) design of the preamp.  Now on to the power amp.  As mentioned earlier, I may eventually develop a range of power amp choices for this amp, but my preferred configuration (and the one I plan to build first) is (2) EL84 per channel.

I've been looking at the circuits of various guitar amps and hifi amps that use push-pull EL84 outputs, to see what I can copy.  One that I especially like, for several reasons, is the Lafayette LA-224.  I have an LA-224A which I've restored: it's our main stereo amp in the living room.  I love the sound, especially with vinyl, though I have to admit the distortion seems pretty high.  I think EL84s are better for guitar amps than for hifis!

I found a schematic online for the LA-224B (various scans of the same original image, at various internet sites); I believe the main difference between the "A" and the "B" is a change in external cosmetic appearance, but maybe there were other changes.  I notice, e.g., that the "B" apparently has dual-ganged pots for bass and treble, i.e., a single knob adjusts L & R together, for each; the "A" has concentric pot shafts so that you can adjust left and right separately (which is usually more annoyance than convenience: I approve of their change here; unfortunately, the cosmetic changes were not an improvement, the "A" looks much better IMO).

This is my re-drawing of just the power portion (one channel):

This combination of tube-rectified power supply, cathode-biased EL84s, and negative feedback, is the basic formula that I'm looking for, so I figure I may as well start from a known-working circuit design.

The Eico HF-81 was another possible template to copy, but that amp has R-C stuff in its feedback loop.  The LA-224 has a nice, simple negative feedback loop (the 7k Ohm and 200 Ohm resistors); it will be easy to add the presence and resonance controls into this circuit.  With the HF-81, it seems to already have frequency-shaping in the NFB loop, and... I dunno, I like it simple.

Really, the only subtlety in the LA-224 schematic above, is the R-C circuit in between the two 12AX7 stages.  From what I can tell, this is a lowpass filter with a very high cutoff frequency, around 2 MHz.  It reduces the gain well above the audio band, presumably to prevent oscillation.  (Maybe the R-C components in the HF-81 NFB loop accomplish this same frequency-compensation task; regardless, I like that the LA-224 doesn't have 'em there.)

relay control section

This is basically the only solid-state circuitry in the amp, other than diodes and LEDs.  Everything will fit on a small circuit board behind the six-switch "sx" panel.

The relay drive circuits use sections of the LM324s as comparators.  A given input goes "low" to turn on the relay, i.e., to enable the circuit section in question.  Inputs can be driven "low" either with the six front panel switches, or using the optional footswitch unit.  The LM324s drive 2N2222 transistors for greater current capacity; the transistors should handle up to 600mA each, more than enough to drive two parallel relay coils (as in the reverb section).

The 1M resistors produce a hysteresis band through positive feedback: i.e., if the input voltage drops to a low enough value to just barely flip the output to "high", it must then rise more than the width of the hysteresis band (perhaps about 0.5v) before the output will flip back to "low".  This prevents oscillation or noise when the input voltage is close to the threshold voltage; the threshold shifts slightly, making the output "reluctant" to change.

This control circuit is completely isolated from the audio portion of the amp (note the use of a different ground symbol).  Power comes from a small plug-in DC power supply (aka "wall wart"), which plugs into an AC outlet provided for this purpose, inside the amp.  This outlet is switched by the "preamp" power switch.

Originally, I thought I could use the stereo effects return jacks as power amp inputs; doing this would have required some tricky wiring in the bypass relay circuits, so that the jacks would function properly even when the preamp power is off.  However, I have since relocated the effects returns such that they can't double as power amp inputs, anyway.  So, no tricky wiring, all six bypass circuits operate the same way.  And there will need to be four additional jacks provided on the back panel: stereo pairs of "pre out" and "pwr in".