Vox AC4/EF86 build/rebuild

[terminalgs]

One of my first ground-up builds was an AC4 built into  one of those modern Vox Pathfinder cabinets with an 8” speaker.   I chose the original EF86 Vox AC4 circuit, and opted not to build it with a tremolo.  I considered it successful at the time, although it wasn’t as loud as I hoped (I knew it wouldn't be loud,,, but I didn't know its be that shy!).  As an attempt to do something about the gain level, I added a 12AX7 to the front of the EF86 in a gain+CF “Top Boost” arrangement with a gain knob between it and the EF86. It was a fun experiment, but the amp, tone wise, and fun-to-play-wise was in the end... blah...

For this reason along with the fact that the gut-shots of my first amp build were somewhat embarrassing,  I decided to gut it and rebuild.  After an elapsed 5-6 years of continued reading, learning, and building, I hoped for a big improvement of the original build.

I'll cut to the chase:

link to schematic: http://i.imgur.com/wKRkzso.jpg



there are some things on that schematic that I'll admit... spiraled out of control.  namely the tone stack(s).  I was *this close* to corking the tone pot hole in the chassis! then I got some ideas about a tone stack experiment and I went with it. (after too much time looking at Ampeg schematics and the hi/low switches).

As built (and in this schematic)  the tone control works well, there isn't any place on the knob swing that is dead, or weakens out.   I'm not sure its any better than, say a Magnatone or Fender one-knob tone stack.  It certainly is more complex (Building it myself might have been the only way for me to learn that).

I'm still playing with a few things: primarily, the volume pot and load of the EF86, and the grid leak resistor R111 of the EL84.

I followed Valve Wizard's advice closely on the design of the EF86, and I'm happy with it.  as is, Mu=40, and its quiet and stable.  I had to raise the load to 2M to get the signal to be stable, and I'm not sure it that is what the EF86 needed, or what happened to band-aid another problem.  I'm pretty sure I can feed the 12AX7s more signal than ~15%, so I'm going to swap in a 500K volume pot and pull R102 (for starters, at least).

C105 can probably go to 250pf to make the treble end of the tone control more trebly but that's later tweaking. same goes for the value of R109.

I originally had a EZ80 in place of the diodes, but the PT buzzed, and among several concerns I had with that was that the EF86 picked up the buzz.  I swapped in the SS rectifier, and the buzz disappeared. regarding the buzz, any thoughts? the 32uf and voltages seems within the specs of the EZ80.  maybe I was over drawing it, or the PT.  but the PT seems fine now, and I don't think its running hot.  Of course, the voltages jumped way up when I went from EZ80 to SS... from 260Vish to 300V.

anyhow, I wanted to post about it.  I'll update it as it get tweaked.  the trem isn't hooked up quite yet, I was waiting until everything else was working (although the oscillator is oscillating).

if any one has any advice or thoughts, I'd love to hear from you.  I'd especially like to hear thoughts on my voltages on the PT and the EL84.

There are couple things that might puzzle the mind when looking at this; namely: R110, which was going to be a .001 cap, or a resistor strapped with a cap,, or... nothing at all (might come out soon).  and R102 was going to be  strapped with a cap, but decided there are enough tone shaping caps after the 12AX7s.

thanks,
Doug


PS: here's a pic of the amp before I started the rebuild:

[SILVERGUN]

Good story and nice looking schematic....
Thanks for sharing.
 

I'm interested to see how it turns out......keep posting

[andyfromdenver]

Hi terminalgs!

I just did an intro post, and in it lamented the sound of my vox ac4...
Like you, I put it in a pathfinder (what an awesomely cheap cab, speaker, and chassis option), but I did an exact of the 1960's schematic.
It's just a weak amp imo, nothing to like about it!  Anemic at anything but volume max, I can hear the ef86s subtle charm in there, but not at an appreciable level, sadly.  I did some mods to the scheme to eek out gain (have a build thread and pic of modified schematic) and a trem depth (which subsequently alters the ef86 bias).  It has mercury mag transformers the OT is the size of 4 quarters stacked together (not really, but close).
Guess what I did with it.....Gave it to a friend for Christmas!!!!!

I did get some guidance on adding a 12ax7, but was fairly discouraged with my waste of money and happy to move on to better projects. 
If you have some breakthroughs, please share! 
That schematic is very creative!  Do you have an EE background?

Andy

[terminalgs]

I just did an intro post, and in it lamented the sound of my vox ac4...

Hi Andy,

Welcome to the forum, and belated welcome to Athens! (I'm a 20yr resident myself,  we probably know some of the same people).

No formal EE training,  lots of reading and soldering. 

So far, this version is much better than the 12AX7>EF86>EL84 version I did the first time I tried to get more umph out of the EF86>EL84.  It has more clean headroom and its louder. It needs a lot more tweaks before I call it done and button it up.   I'll post a follow up, good, bad, or otherwise...

I don't know if its of much interest to others, but I'll definitely post about the neon oscillator.  I've done a little bit of bread-board type tinkering with these and I want to see if it'll work in an amp in this fashion.   I like the idea of a smaller, cheaper oscillator than the typical triode phase-shift tube style oscillator.  We'll see how it does. 

[andyfromdenver]

Wow, what a coincidence!
Here is a pic of the modded scheme.  The tone cut was an exercise in futility.
Steve Hunter helped me, and tubeswell too (I see he's on here, hi!)

*I hope this uploads...otherwise, it's going to be difficult to share wiring porn etc. from my phone*
*edit, crud looks like i'll have to pull a few shenanigans to reduce the size like email to self etc. that's cool*

[terminalgs]

very cool,  Steve is great.

[terminalgs]

Here's where I'm at after some fine tuning of the tone components.  I changed the P1 volume control to a 500K pot and eliminated R102 250K.

R106 was 220K.  I was getting some high frequency noise coming though when P2 was turned in the direction of R106, it disappeared once the P2 wiper moved to about 150K,  so I upped R106 to the next highest value I had on hand above 370K, which was 470K (this eliminated the noise issues).

I bumped R109 to 180K to make the bass side of the tone control drop later in the knob travel.

At this point, the tone control functions nicely and both ends of the tone control sweep are distinct bass and treble positions.  between about 10 o'clock and 2 o'clock, there isn't much variance in tone, but that sometimes happens with single knob tone controls I suppose.

With tone @ 12 noon, as I turn up the volume, I get clean up to 2 o'clock, and overdrive from there on (with o-scope, I can see the signal clipping en route to the grid of the EL84, so its the 12AX7s being overdriven.

BTW,  original goals are close: It is loud! (a lot louder than previous versions I'd tried) and it sounds really good!

[kagliostro]

Interesting project

can you post the whole last version of the schematic ?

K

[terminalgs]

http://i.imgur.com/0YdOU22.jpg



here's the latest with the above mentioned tweaks. 

like I mentioned before, the EF86 stage is all Merlin, the EL84 and power supply are fairly standard with some influence from past EL84 related posts here as well as some Merlin & Aiken.

The only thing I dreamt up was the treble/bass arrangement mixed together. (the bass arrangement is ampeg's 'ultra-low').  If that looks too adventurous, running a simple voltage divider out of the 12AX7 into the EL84 would work great, I'm sure.

the neon oscillator isn't anything new.  I've got it oscillating, but I haven't tied the LDR to the circuit yet.  I've breadboarded it with the LED and LDR, it works well as a vibrato,, its a big more square wave, where as the the typical phase shift triode osc. is a bit more sinusoidal. putting an LDR directly on the neon is more sinusoidal than the LED, but at faster rates, the lamp tends to not have enough "dark" cycle to get a nice rise and fall variation.  The neon is capable of oscillating very slowly FWIW.

Merlin suggests putting the 12AX7 in front of the EF86 and overdriving the pentode.  I think that'd be great to try.  I tried it the other way around and I think it works great.  I'll try to get some sound clips up after the new year.

[kagliostro]

Thanks for the new schematic

I'm following with interest your development of the little AC4

---

I see you are (as I try to be) a Merlin Follower

so I would like to know if you tried the Morphing Pentode/Triode circuit

on the ef86 tube

---

Tubenit and Geezer has some example of a triode followed by a penthode

but usually they use 5879 (and 5654 if I'm not wrong) the 5879 has a lower gain factor

respect to the ef86 tube but it will be interesting also to use the ef86 setted for less gain

K

[kagliostro]

Here is the Version 3 of the modified AC30 by Terminalgs





I found it this morning searching for ideas about the conversion of the UHER 502 Tape recorder

K

[tubeswell]

I did a goosed up version a while back (that I dubbed 'Super AC4' - call it what you will)  for 1 or 2 x 6V6  or 1 6L6, and it had a morph control on the EF86, a tilt control for the tone, as well as a TS-recovery/driver stage, and a SF buffer for the LFO. It also had switchable HT windings and could take a range of rectifier tubes. These SE amps sound best with a 10" speaker IMHO.


https://www.flickr.com/photos/75190241@N07/8031948687/in/photostream#

[77barius]

I know I'm reviving a lost since buried post but I blew out my AC4C1 and decided this is the version I'd like to rewire. Think I have my layout sorted, but I'm not sure if you are tying the grids of the 12AX7 together.

Can you advise please? TIA

[sluckey]

I'm not sure if you are tying the grids of the 12AX7 together.

They are tied together.