AA864 Build and ? switchable overdrive stage

[Bill Brown]

Hello,

I am new to this board, but have read a lot, and y'all seem like the folks to ask about my project.  I hope you will indulge me.

My 20 y/o son plays a Gibson Les Paul with P-90's.  His band plays "Texas country," blues (he would love to play more of the latter, worships all things Duane Allman).  He has a Gries 5 that has been great for practicing, etc.  He needs an amp for gigging in small-medium bars (has been using whatever he can scrounge).  For cleans/rhythm, he likes Fender.  My idea was to build a ~45W Fender circuit in a "head," plugging this into the Gries 5 speaker, probably eventually a 2x12" cabinet.  This would hopefully "take pedals well" for lead.

All of my reading leads me to the AA864.  I re-drew the circuit (helped me to work through it) and have all the parts here or on the way to do everything on it.  I tried to be complete; every wire/connection that I can think of is on there, it is what I would build from.

The only changes I made (that aren't accidental!) are using only the normal channel, tube rectification (56FA style), and adjusting the last dropping resistor for the lack of the other small signal tubes.

Do you see any mistakes?
Are there not grid stoppers on the preamp and PI stages (on hi-fi schematics they are assumed there, but it seems not on guitar circuits)?
The grounding I have tried to convey is modified star.

I have a circuit that I evolved from this that has possible modifications, including an added/switchable overdrive stage, but this is quite lengthy so I don't know if I should include it now.

Thank you for any thoughts,

Bill

[Bill Brown]

Oh, and I also used a multi-tapped output transformer, taking the NFB from the 4 ohm tap as original.

Bill

[sluckey]

Your circuit looks good to me. You don't need grid stoppers in the places you asked about. You also don't need the cap across the power switch.

[eleventeen]

At the risk of discouraging you from a neat project that I know you want to build for your son....don't you think you could buy a decent condition used Peavey amp for about half of what that would cost you to build?


You think you can build this for $200? I see these for sale all the time in the low $200's.


With reverb, channel switching (this one has a better speaker which is the big weakness over the stock item) cathode bias so you don't have to bias it.


I have one of these, swapped an Altec speaker into it. It's 90%+ as good as my old 40-watt modded SF Princeton Reverb. They will sound better with better-than-Chinese (stock) tubes installed but there is nothing wrong with it as it is.


 

[Bill Brown]

Thank you,  I have read lots of your good advice here.

I wondered about grid stoppers from two perspectives: Filtering RF, but also to help with "blocking" from grid current.  Coming from hi-fi amplifiers where the gain structure is closely specified and you always bias to avoid grid current, it seems these high gain circuits are more likely to experience it.

It is very difficult to grasp the overload/distorting mechanisms, etc. that lead to the (desired) sounds from these amps!

Thanks again,

Bill

[Bill Brown]

Eleventeen,

There is no doubt that I can't build it for anything near that price.  I suspect the transformers alone were close to that :).

I have really enjoyed researching these old amplifiers/schematics, learning, designing, planning, etc.  Spent a lot of the weekend scheming on the cabinet design with my son (he is turning into a decent daytime carpenter/nighttime musician(!)).  Will hopefully work on it with him, my father and brother over Thanksgiving.

If we end up with a working amplifier I think it will be worth it.

Bill

[eleventeen]

I appreciate that it might be a really fun project for all and I am glad you take my suggestion in proper light. It's just that the finished U-build-it project requires the simultaneous exercise of woodworking and metalworking and obviously electronics and that fairly big check for all those parts....and a handle and corners....and $20 worth of knobs..10% of the whole amplifier in just knobs!!..and guess what, if it doesn't quite suit his needs, he can probably sell it for about $200. Meanwhile, yeah, it's made in China but it might just be a lot more practical in terms of a rugged thing he can chuck into the trunk of his car and have lots more survivability than something you might build. I have mentioned these amps a few times and I apologize if I seem obsessed over them; it's just a ridiculously practical cheap thing that has worked very well for me, they are built very well and with a good speaker they sound quite good. The stock speaker is simply a travesty! But it makes a good bench speaker once you swap it out)

OK, I'll shut up now.

[Bill Brown]

It's ok; I do understand and appreciate your thoughts, and suspect there is wisdom there,

Bill

[John]

Financial wise, 11teen is right. But I can tell you, hearing a working musician playing through something you custom built with a smile on his face, is priceless.

[Ed_Chambley]

Hey Bill,
The AA864 is a very nice sounding Bassman.  My second Favorite.  I am really liking BrownFace amps right now and play a 6G6B through 2, 12 with Altec speakers.  Interesting you are using a Tube Rectifier, but it will be cool.

I am interested in seeing how you plan to implement an overdrive.  Do you have that drawn and mind sharing?

[sluckey]

...is priceless.

Especially when he brags to his friends that my paw and grandpaw helped me build this!

[Bill Brown]

Thanks for the ongoing thoughts, guys.

As requested I'll attach the modified schematic I came up with.

I started scheming in this direction as I was becoming apprehensive about him putting a cheap overdrive pedal in front of the amp to use for soloing.  The list of possible gain-boosting ideas I found included lifting ground on the tone stack, adding a switchable gain stage, increasing input resistance to first stage, increased anode load resistors, elininating the voltage divider resistor into the PI stage, changing the value of the NFB resistor or eliminating NFB, etc.

I changed the schematic to show some of these changes, not all (and also with a midrange pot v fixed R, and how it might look with a cathode-biased output stage) and circled them in red.  (It still has for now the early stage grid-stoppers and AC filter cap that I was told could be eliminated).

The boost stage relay/PS isn't fully complete, it is obviously modeled on the work done on this site.  I figured I could use about 3.6mA through the stage to make it fairly stout having not included the small tubes from the bass channel in the original.  With this and the available voltage I then just calculated a potential operating point.

Having a day or two to think about it, it might be wise to build it in basic form first, and maybe focus on a good boost/overdrive pedal.  I just don't know enough about magnitude of overdrive gain needed, where in the circuit to place it, etc. 

Certainly any thoughts would be welcomed,

Bill

[sluckey]

You should remove the 470K from the output of the boost circuit. Otherwise it will upset the delicate bootstrapped bias of the PI.

[Bill Brown]

Ahhh..yes, of course, makes sense.  Thank you.

I wonder, then, about charging/discharging currents on the capacitor, and then wonder if the sequence of gain stages is correct (where in the sequence the overdrive is placed).

Bill

[2deaf]

The 500pf coming off of V1b is fine when it sees the large impedance of the PI, but it may be a little skinny when coupling to the overdrive stage.

The overdrive level control won't perform its intended function.

[HotBluePlates]

... I was becoming apprehensive about him putting a cheap overdrive pedal in front of the amp to use for soloing.  ... Having a day or two to think about it, it might be wise to build it in basic form first, and maybe focus on a good boost/overdrive pedal.  I just don't know enough about magnitude of overdrive gain needed, where in the circuit to place it, etc.

I think the practical challenge is balancing your distorted volume vs. clean volume. Getting an amp to distort is easy; getting it to distort on command by not just "making it louder" might be a challenge.

I bring the issue up because it depends on your son's playing and the band's repertoire. Required amp power for live playing is pretty much all about how loud you need the clean sections of the song to be. I think you'll find most of the methods of boosting gain do that, but increase volume as well. That's good for a solo boost, but not if he's called upon to switch clean to distortion & back at largely consistent volume. The evolution of preamp distortion/master volume amps seems to be about getting switchable distortion in the amp at a volume level consistent with the clean output of the amp.

So that's what pedals offer conveniently: switch from clean to distorted on any amp, with many flavors of distortion available. Back in the 80's a friend of the family was a semi-pro guitarist playing in a Top-40 band. His rig to cover 90% of music out there was a Les Paul, a Twin Reverb (clean as loud as needed for any room), a DynaComp compressor, a RAT distortion, a Boss CE-1 for tremolo/vibrato, a delay pedal (probably an old MXR delay) and an MXR Phase 90. He may have had a wah & volume pedal too, but I don't remember right now (the volume pedal seems very likely, the wah less so).

So the route he took was lug an amp with all the clean output power you'll ever need for any gig, while pedals offer all the effects and distortion at any volume without needing to lug a separate amp or flip any switches beyond footswitches.

I mention it for your consideration, and by way of suggesting the suitable amp design is one which solves his practical problems and suits what/how he plays most of the time right now.

[2deaf]

The cathode resistor on the 6L6's is way too large.  If you put something like a 250 ohm resistor in there, it needs to be at least 10W.

The evolution of preamp distortion/master volume amps seems to be about getting switchable distortion in the amp at a volume level consistent with the clean output of the amp.

To complicate it even more, it seems like no two people can agree on what exactly is the same volume level between the two signal types.

[Bill Brown]

Aha, this might be helping me to understand the crux of the matter.

With the low powered amp he has great tone and can find the correct volume for playing lead.  He can't get good clean to play rhythm.  I told him, mic the small amp and use the gain pedal-switch or get more watts.

When playing live he needs a way to play rhythm, then get a little louder to stand out of the mix and a little to medium more dirty.  He will never play with a fuzzy or metal-type sound (well, never say never, of course).  He pushes his Texas country band toward a Southern blues sound with his style and the sound he wants- so far they are ok with it, though they will cut him off if he gets too "Mountain Jam-ish" :).  He loves the Fender cleans, but wants to be able to kick on what he describes and hears in his head as a "growly scream."

So then is what is needed a pedal that pre-distorts the sound a bit and moves the amp toward clipping without a marked increase in the loudness that you hear out of the speaker, i.e. the Bassman and a (?) Tubescreamer?  Is that what it would do?

Bill

[Bill Brown]

thank you 2deaf.  Before calculating/re-sizing the resistor, is there a consensus (ha!) on cathode v. fixed-bias in this setting?  Having read a fair bit, my temptation is to stick with the original schematic, fixed bias.

Bill

[HotBluePlates]

With the low powered amp he has great tone and can find the correct volume for playing lead.  He can't get good clean to play rhythm.  I told him, mic the small amp and use the gain pedal-switch or get more watts.

When playing live he needs a way to play rhythm, then get a little louder to stand out of the mix and a little to medium more dirty. ... So then is what is needed a pedal that pre-distorts the sound a bit and moves the amp toward clipping without a marked increase in the loudness that you hear out of the speaker, i.e. the Bassman and a (?) Tubescreamer?  Is that what it would do?

So you have a couple possible solutions there. It sounds like if nothing else he needs a large-ish Fender amp for cleans, so you can pencil that in as a definite.

I once modified a blackface AB165 Bassman several ways, but the main one was to exchange the existing Bass channel for a 5F6-A Bassman circuit (the amp has an unused triode, so I was able to have both the blackface preamp and the 5F6-A preamp). I used an A/B/Y pedal to select the channels. At the same Volume knob numerical setting, the 5F6-A had much more mids and volume/gain. If I turned up the blackface channel to match volume-in-the-room, then both channels had the same amount of distortion (though the sounds were still distinctly different).

The lesson for me in that experiment was that in the old Fender amp circuits without master volumes, the first thing to distort will be the output tubes. That happens when you feed a large enough signal to them no matter how you derived that signal.

Thought Experiment:
What if you had 2 complete channels with no shared components before the phase inverter, one with your existing 2 stages and tone circuit for clean, the other with 3 stages and its tone circuit for distortion. Assuming the 3-stage channel develops large enough signals to cause the preamp tubes to distort, you could balance the volumes of the clean and distortion paths by having a level control for each right before the phase inverter. The distortion channel level could be brought down to the same level (or slightly more, for  boost) as the clean channel. The phase inverter and output stage cleanly amplify what's presented to them to drive the speaker.

Your present schematic is a bit more elegant, in that you're adding or removing a gain stage for the distorted channel. If your switching also included some form of level control for the output of each path ahead of the phase inverter, you might be able to balance the channel volumes.

That said, folks who modify Fender-style clean channels for distortion also typically reduce the value of the cathode bypass caps from 25uF to maybe 1-5uF because distorted bass gets muddy fast. And the tone control settings for clean may not be the ideal tone settings for distortion.

For an example preamp using this concept but sharing some gain stages, look at page 2 of this schematic. The project uses London Power's preamp, with 2 gain stages for the clean path, and an additional 2 gain stages for the distortion path. Each has individual level controls. The output ("Jack-O") can go straight into your phase inverter and output stage of choice. Switch 4 (Channel Select, in the middle of the schematic) could be accomplished with a relay, front panel switch, or solid-state switching circuitry; however, I would not advise running the circuit out directly to a footswitch DPDT because it would like cause hum & noise problems. Keep the switching element right at the circuit and run wires from control elements to a footswitch.

Now I'm not saying this is the best preamp or solution, but wanted to give some alternatives and get you thinking about what exactly is the practical problem you wish to solve.

[Bill Brown]

HBP,

Thank you for your nicely-detailed, long note.

I went scurrying to find my printed version of the 5F6-A.  It is actually the Bassman circuit that I researched first, learning a lot by reading through the circuit analysis on the "Amp Books" site.  I went with the AA864 ultimately for the blackface v tweed clean.

I had read about people converting the bass channel of these amps to a different preamp circuit to have two guitar channels (tweed + blackface, guess you could have blackface + JTM45, etc.) but was (initially anyway!) shooting for simplicity.  It was only when I thought of using a cruddy pedal in front of the amp that I started thinking of an overdrive stage, found the relay switching on this site, etc.....and went down this rabbit hole :).

The beauty of building the blackface clean and tweed dirty (which I think is what you were ultimately describing, or how it would be used in practice) is that the gain structure/operating points/etc. are established, copyable v. trying to shoehorn in a gain stage and having to experiment/tweak to get it right.

So, scrap my second circuit :).  I have the parts (or they are on their way) to build the normal channel of the AA864.  So it is either that alone (with pedal overdrive/boost), or the AA864 preamp stage and the 5F6-A preamp stage (normal channel) into the AA864 PI/output stage.

The complexity of the assembly can be managed (with a bit of trepidation, granted!), the question is the switching.  Is there an elegant, reliable way to switch between the two inputs?  I looked briefly at the A/B/Y boxes after reading your post and will resume the quest in the morning.

Thanks again,

Bill

[PRR]

Blackface is an excellent choice for this (and much) music, and AA864 is a fine Blackface.

Many of us who lived through the 1960s "been and done the AA864", which may be why Ed is exploring other eras. But as a First Car, you put your child in a Volvo/Subaru/Toyota. If he ends up with a Masaratti or Willys or a Fit, that's his path to find over time. The 864 was the Impala of its day, which today is a Camry.

5F6 is another classic, somewhat simpler. IMHO it has a narrower repertoire than the 864, but of course it has many fans.


Blocking resistors are generally newer than guitar amps.

There's some discussion from Williamson about blocking resistors to minimize transient in-fidelity in what became THE Hi-Fi amplifier; but the fad for series grid resistors everywhere is a fairly new thing.

Fender input jack standard wiring includes a 68K/34K series resistor which (with grid capacitance) helps roll-off the AM band. Once you get that far, inside a metal box, you probably won't have more problems.

Many Hi-Fi designs use tubes at very high current, for supersonic treble extension, which makes hypersonic self-oscillation possible (and by Murphy's, probable), especially with the sometimes roundabout layouts used in modern Hi-Fi. Working with 220K plate loads you just can't get much supersonic; Fender's 100K loads suggest *close* study of Leo Fender's layouts which are sometimes just-stable and may need a little pencil-poke to move wires to less-unstable positions.


> over Thanksgiving.

I built most of a Champ chassis, in stolen hours at work, and it was a loooong job. Yes, I was doing "different" which meant re-doing some original work a couple times (be ready for that). I was having fun working, you guys may be more goal-oriented. But a 864 is no snap.

If he has a X-Mas gig, I would keep his eye open for the $200 Pea-Heart 113 (several similar ubiquitous beasts) in Craig's List. Good examples don't suck, speaker is often limiting yet easily replaced, and it can always be the backup for the "good amp".

[Ed_Chambley]

PRR nailed it for me.  I have played BlackFace fenders for years and years.  Don't get me wrong, I love them and it is necessary for me to have that available.

My Bassman is a original so I would not drill a hole for switching.  It is the reason a lot of guys use A/B/Y boxes, but if I were scratch building I would make the switching internal.  The Bass channel on mine is (was) funky for guitar.  Made a couple of changes to achieve a boost, but not necessarily looking for distortion.

IMO, if you son plays venues where he can get 40 to 50 watts cranking, it will not be hard.

I hate to say it, but all the professional players I know use pedals.  There will always be the guys who do not, but I just don't know them.    I will recommend a effects loop to use for Modualtion.

Personally I always play with 2 amps.  Hey, I carrying a backup so might as well use it.  My main amp, the bassman channel switches and the A/B box I built has the ability to balance the signal.  I am sure HBP suggestion of internal balance would work best.

I have my main tone and I put overdrives, distortion, compressors and fuzz pedals (which BTW I use to get a vocal note tone).  The other amp is my wet amp, usually small and easy to carry.  I use a 2, 12 in stereo if it is a decent size place.  All the Verb, Delay, Echo, Chorus, Phaser, Flanger go to the smaller amp.  This one changes from time to time according to where I play.

I have known people attempting to get that SRV tone.  Takes a lot of amps running at the same time.

Just my 2 cents.  I was mainly saying I have never been happy gigging with one amp.

The biggest thing is having the time and interest and support in your son's playing.  Build him a solid roadworthy kickin amp.  He will never forget it.  When my Father gave me his 64 Super Reverb and 69 Black Beauty Les Paul, both never gigged it was the coolest thing and I will never forget.

[Bill Brown]

Thank you Ed and PRR, more wise advice and other interesting thoughts.

What you say about grid resistors makes sense.  The one on the input should decrease RFI getting in to the chassis, then everything else is inside the box.  I have read that the Williamson, with its high gain for high NFB was always on the edge of stability so they were doing everything they could to manage bandwidth/phase margin, etc.

With some high-gain, high-transconductance tubes in Hifi (6C45 is fairly notorious in this regard), people end up with stoppers on every pin to keep them from oscillating.

In terms of "blocking," I guess the output tubes in these circuits are most susceptible, so there you find grid resistors.  This also goes along with HBP's suggestion that in these circuits the output tubes are the first to distort (when I first investigated guitar amps it was suggested that this was the most desirable re. tone, but it seems like the general emphasis these days is on preamp distortion- perhaps because it can happen at lower volumes(?)).

I suspect that no matter what I do he'll be sticking pedals in front of it- hey even Duane Allman (his hero, well, actually more than that) used one ("with a run-down 9V battery"), so I guess my best bet is to stay simple, build the AA864, and encourage him to find (or even build) good quality ones.

You guys have really helped!  Thank you,

Bill

[Bill Brown]

And if he had that '64 Super Reverb and '69 Les Paul!  Wow!  He would be done!  Awesome.

[PRR]

> Williamson, with its high gain for high NFB was always on the edge of stability so they were doing everything they could to manage bandwidth/phase margin, etc.

Not that complicated. Overdriven transients cause distortion to subsequent waves via blocking. Instead of a 1mS clip, you get a 50mS smear. For Williamson (armed with shiny new oscilloscope), this was A Problem. For ('scope-less?) Fender, it wasn't all bad, and sometimes good.

The 5K resistors at Fender 6L6 grids *appear* to be to limit treble extension (the pentode has more extended treble than a triode because of less Miller effect). This means the 6L6 plate leads are less likely to sing-back into earlier stages.

Williamson used 50+K resistors, which did cut into the extended treble needed for heavy NFB. However if the coupling cap only charges through 50K, discharges thru 100K, an overdrive causes very little bias-shift and grid blocking. Looks better on the bench 'scope. Can not "grind".

IAC.... you are not asking about some odd unpopular and unloved amp. The BlackFaces (and many-many copycats) are THE sound of Rock. Rock goes under many names and styles, Jazz cello through a Twin, that stuff they play in Texas, stompin' church music. So whatever you may think about the theory, your first path should be to copy the classic verbatim.


Whether you scratch-build or buy+mod is of course not a rational choice. The get-er-done path is to admit that Chinese women can do the work and get it to market in bulk far faster and cheaper than you can ship a couple heavy parts. I don't grow my own cows and coffee-- folks do that better than I ever could, and conversely I could do better managing computers for money for cowburger and beans. I even buy whole frozen cheeburgers, even if I "mod" them with better cheese and fresh tomato.


What I guess I'm saying is... a two-6L6 amp is mostly used where people are "drunk". Roadhouses, bars, etc. (And in a very different venue, churches where the spirit isn't Jim Beam but is again enrapturing.) In My Opinion, in such work you do not need a "magic" amplifier. After the third beer (or Amen), most listeners are only hearing the surface of the sound. Any well-behaved amp of reasonable EQ, speaker throw, power, and non-annoying overdrive (plus pedals) gets the same tips (tithes), applause, and call-backs.

There is the issue that "I play better with a better amp". Such self-assessment is hard to analyze. Arguably a professional (even if only part-time) "should" be able to do his work with any tool which does not suck, whatever tool best balances the music and the expenses (labor).

I'd save the craftsmanship for a smaller amp more suited to summer afternoons with friends and family, where the music is SO good that the beer is just throat-wet and goes flat before finishing. A big Champ or a DeLuxe. The big BlackFaces were the workhorses of their day, but today IMO the transistor amps do most larger venues lighter and more reliably (and far cheaper, which is an issue when an amp may be destroyed by party-animals).

[Bill Brown]

Great post.  Anyone that can combine Chinese women factory workers, Jazz cello, Jim Beam, stompin' church music, and make an analogy between a third beer and the "amens" of a southern church into one forum post is someone I'd like to have all three beers with.  You are pretty far north to sound like somebody from Texas :).

Other than vacuum tube rectification and replacing the mid resistor with a pot, verbatim it is.

Your nice description of expertise, paying for it, and sticking to one's own is also understood. 

I have been a volunteer for the last 2 years in Southern Belize.  I am in the States for a few months away from my wife and youngest son working (was running out of money!); I have time in the evening so it will be a good distraction.  Plus, I can't back out now, the parts are bought :).

Bill

[Bill Brown]

So, over the Thanksgiving weekend we (my father, brother, and son) were able to build the cabinet and finish the metalwork I had started.

I then left the country for four months, returning three weeks ago (my son has been playing mostly acoustic, Gries 5 for practice, borrowed amp for gigs).  Since then I have painted the box, installed the hardware, and assembled the components in the amp.  Yesterday evening I went back through every connection referencing the schematic.

This afternoon (with a bit of trepidation) I began live testing and followed this sequence:

1.  No tubes, power switch on, plug it in- pilot lit, fuse ok, proper voltage at the pilot, left on for 5 or so minutes, no smoke/heat, unplugged
2.  Rectifier in, plugged in- fuse ok, ~464 volts at the top of the first filter cap(s), unplugged
3.  Added preamp tubes- good (I think), decreasing DC voltage on each filter cap, bias voltage present at the cathode of each preamp tube (though they seemed a bit high), adjusted the bias potentiometer to maximum negative voltage
4.  plugged in power tubes, plugged in cheap 4 ohm car speaker to 4 ohm jack, plugged in power- everything seemed ok, a little noise coming from the speaker, don't think I heard any hum.  I then tried to adjust the bias pot up, aiming for -44V.  Loud squeal, unplugged.

I tried to follow the schematic to end up with the proper polarity at the OT.  I am hoping this is just that I am injecting positive feedback into the cathode of the phase splitter and there is an easy fix(?).  If so, is the most straightforward fix to reverse the 6L6 anodes into the OT, or (please say there isn't) a different problem.

I have attached the final schematic.

Thank you for any advice,

Bill

[Bill Brown]

yep, disconnected the FB and squeal gone.

[sluckey]

If so, is the most straightforward fix to reverse the 6L6 anodes into the OT,

yes

[Bill Brown]

That did it :)

The hiss was so much less I wasn't sure it was on.

[Bill Brown]

Here are some pictures, got everything put together this morning.

With this my son should have available 5 watts with the tube amp built into his current amp/speaker cabinet and 22 or 45 watts (with a tube change and re-biasing) by plugging this new amp into the speaker. Should cover everything from practicing to recording to small gigs to larger gigs.

Thank you very much for the advice and wisdom,

Bill

[Bill Brown]

pic 2

[Bill Brown]

pic 3