Eric Barbour's 6BM8 One-Tube Reverb Imagined...

[dchang0]

Since I can't find the schematic for Eric Barbour's famed one-tube 6BM8 reverb for Fender amps, I decided to take his example (Fig. 2 from this article: Tube-Town Germany - Hot Stuff Cool Sounds) and flip it around backwards to form a complete one-tube reverb circuit.

BUT, I don't know what I'm doing (been a while since I did this sorta thing). So, I need your help double-checking the schematic, especially since I don't understand grid bias, etc.

The original schem has a 180K resistor off of B+ which eventually feeds the control grid on pin 3 of the 6BM8. I plan on using the 5F1 champ's first preamp stage, which has a 100K resistor there instead.

Here's Eric's original schematic:



And here's my imagining of his reverb circuit that needs double-checking:



Thanks in advance for any help you can provide!

[überfuzz]

I don't know what reverb tank you intend to use, but you should consider the impedance of it. Likewise, when you flip around the stages of the reverb circuit like that, be mindful of the impedance.

[dchang0]

It'll be the Accutronics 8EB2C1B, with 800 ohm input and 2575 ohm output. It's a shorty tank (this is going into a 5F1 Champ cabinet).

[überfuzz]

Ehh, did you say you wanted to mod a 5F1..? Did you ever consider leaving it as it is and get a fender amp with a stock on-board reverb. Off topic, but still...

Right. Why don't go for a Fender reverb circuit. Most of the on-board reverb circuits by fender are the same. You'll have a fender amp, with a fender reverb. Yey!!! I think you'll be able to use you're tank to. You might have to adjust some values to make if work nicely.

[dchang0]

Heh, heh, heh--'cause I like sitting up till 2AM reading about tube amp design?

LOL. All kidding aside, there is a real reason for wanting a 6BM8 one-tube reverb. I'm using a real 5F1 cabinet and chassis, which are tiny (and into which I can only fit one extra 9-pin tube). So copying any true Fender reverb circuit is out, as they tend to use multiple tubes for their reverb.

As for why a 5F1--space and especially wattage considerations dictate the small size. I am actually selling off my only-slightly-larger Princeton Reverb build (which has a great reverb) and keeping the 5F1 build. The key is: I don't PO the gf (space) or the neighbors (wattage)...

And plus it actually IS fun to go back and re-learn about tube circuits. It was almost 20 years ago when I studied this stuff (which is why I need help spotting any errors...)

[jmaf]

My 2 cents: You probably need a grid leak on the triode, 330k to 1M from grid to ground. The reverb output says reverb send twice, just a syntax error there.

Don't tie the reverb send ground to chassis ground as depicted, just run the 2 transformer leads straight to the tank isolated. Ground to chassis only the tank return ground.

.22 return cap seems too big. 0.022 more like it. Even smaller will do, my last one was in fact 1nF

2M return pot seems way too much as well, my last project used 250k because you want the return gain lower than send, it's very noisy on the single coil pickup at that tank.

Hope these ideais help somehow.

[BiBi]

+1 on what JMAF says, although the return cap probably depends upon the how much signal you have filtered at the input.
You'll also need to look at the amount of drive you need for the pentode section of the 6BM8. It's not a small bottle EL84 regardless of what you may read; it's more like a 6AQ5 or small bottle 6V6. Secondly, the tank that you have isn't really optimal for the transformer output. You might have better luck by making the pentode a triode which still offers plenty of output and makes the circuit less picky about impedance.
You might also look into using a transistor recovery stage-- a JFet/Mosfet cascode circuit works really well for this, probably better than a single tube.
Merlin's site The Valve Wizardhas some interesting circuits for medium impedance tanks as well.

[dchang0]

Thanks, guys!

Okay, I updated the schem with jmaf's suggestions (all good ones). It is currently at Rev. 1.0a.

If it helps to figure out the proper return cap value, the input stage is the very common, standard 12AX7 triode first gain stage on the 5F1, virtually identical to the triode stage of the 6BM8 as shown (100K plate resistor instead of 180K, 22uF cathode cap instead of 47uF, all other components the same values).

@BiBi--it's funny, I happened to be reading The Valve Wizard's article on reverb drivers last night at 1AM! Eye opening for sure... I'll have to go re-read it in detail to get a grasp on the math involved.

Using a triode for the driver section is definitely an option. Others have said that a 12AX7 is too weak to drive a lush reverb circuit. I chose the 6BM8 mainly because of all the good results people have been getting with them (surprisingly few schems out there, though). Plus, to some degree, figuring out the 6BM8 is a learning experience for me, as there are plenty of "plug-n-play" one-tube-12AX7 reverb circuits out there that I could simply copy without thinking. (I still may copy them simply because it appears that the number of components can be reduced by one or two, saving space.)

As for the lack of an optimal tank, I happen to have on hand a Heyboer reverb transformer (25000 ohms red-blue, 8 ohms green-black) that's a drop-in replacement for the Fender reverb driver. If need be, I could buy another, optimal transformer and sell this one.

I haven't yet ordered any reverb tank and so could easily choose a better one. HOWEVER, it doesn't appear that tubesandmore.com has anything higher than the 800 ohm input model in the Type 8 Accutronics, though I believe Accutronics does make much higher impedance models. I have no idea where to buy those.

[BiBi]

you want a lower impedance tank with transformer coupling. I think the "A" model is the most common with 10ohm input.
That said, your transformer is for a typical Fender style circuit with a 12AU7, thus the high primary impedance. I believe the single ended 6BM8 pentode OT would be more in the 6-8k range.
I personally like the Ampeg style of capacitor coupling but then I'm a cheapskate!

[dchang0]

Oops. I thought you were referring to the later paragraph in The Valve Wizard's article where he says "Cheap current source drivers:
Low impedance reverb tanks seem to be persisting in valve amp design, which makes little sense since valves are not well suited to driving low impedance loads. By taking advantage of high-impedance tanks (usually intended for op-amp drivers) we can build much simpler valve drivers." The example circuit he gives with paralleled ECC82s is nice in that there's no reverb transformer, but I don't have room for two tubes. Maybe if I could pull it off with the pentode side of the 6BM8....

He does suggest low-impedance tanks in the paragraphs before that (where, like you pointed out, he suggests using a triode instead of a pentode).

Which do you think would be better-sounding? I'm at that fork in the road where I could go either way. I LOVE lush reverb (got to, if I'm crazy enough to try to stuff an entire tube reverb circuit into the tiny 5F1) and frequently crank it up to surf levels (though I don't play surf guitar).

The goal is to have the reverb be as true a representation, tonally, as the original signal.


The Ampeg method is definitely good for saving both $$$ and physical space. I might try that, though most who have used it say that it's nowhere near the Fender surf sound...


P.S. All this raises a new question: Would it be wiser to use the triode side of the 6BM8 to drive a low-impedance tank, then use the pentode side as the recovery section? Would I be able to get the surf sound outta that, do you think? The Valve Wizard warns against using the 12AX7 and suggests the 12AT7 or 12AU7; I'd heard that the triode side of the 6BM8 is a lot like the 12AX7, so this might not be such a great idea...

[BiBi]

Well first off the Fender surf sound I believe is from the stand alone unit. I haven't heard any of the amps that sound that crazy! I think when you put the springs into a combo amp bad things can happen. You always have to make compromises such as more filtering that you may not have to in the stand alone. As well, the stand alone unit uses three tubes I believe.
You could do something in that vein with a 6BM8 and a second tube pretty easily. Just don't give short shrift to the recovery section as that's just as important and maybe somewhat more difficult. I would think of say a 12AX7 as the input, volume control into the triode-connected pentode section of the 6BM8 capacitor coupled into a high impedance tank like the 8F.... Then use the triode section of the 6BM8 as the first section of recovery through a tone control, volume and into the second side of the 12AX7.
As you'll need a "dry" signal as well, try a source follower transistor setup like a mosfet irf820.
That's just my 2 cents. I have done something similar with a EL84, 12AT7 and a 12AX7. I could get a really wide arrange of sounds especially with the tone control.
The unit is now being used in a music studio.

[dchang0]

LOL--crazy is right! Well, there are definitely a ton of constraints here that will force a serious compromise on the sound, so while "aiming for the surf sound" is cool, actually getting the surf sound is highly unlikely.

I like your idea, though there just isn't enough physical room for the circuitry. Here's what I'm thinking of doing:

Guitar input -> 1st 1/2 12AX7 -> split off reverb from dry -> reinsert reverb -> volume pot -> 2nd 1/2 12AX7 -> 6V6.

Reverb: [-> pentode of 6BM8 -> tranny -> tank -> triode of 6BM8 -^]

I like your idea of doing a straight capacitor coupling into the tank:

Reverb: [-> pentode of 6BM8 in triode wiring -> tank -> triode of 6BM8 -^]


I only have space for the one volume pot on top plus stealing one of the instrument jacks to install the reverb pot. (The problem is that there's only 6.5" of control panel on top, and the chassis is predrilled in the standard 5F1 layout, so there's no room to fit another pot.)

[BiBi]

Oh I should actually read what you wrote at the beginning-- I had no idea you were trying to fit this into something so small. The problem is that when things are packed in really tight, spring transducers tend to feedback and fighting that is absolutely no fun. If I had to build a reverb in something like that I'd probably go for the Belton digi-log. You can also drive these with a single tube and they sound pretty darn good I think. Maybe you should save your parts for a stand alone unit-- the general consensus is that reverb with SE amps doesn't go so well as you generally want a really clean sound to go with reverb.

[dchang0]

No biggie. Really, I shouldn't be talking too much about the 5F1 idea because this thread is meant to help the DIY community regarding the hard-to-find Eric Barbour 6BM8 mod to Fender amps.

I am planning on starting another thread with the 6BM8 circuit actually inserted into the 5F1 after I figure out a pretty good (generic) approximation of the Eric Barbour circuit. So my own preferences (surf sound) and constraints (tiny chassis and cabinet) don't really matter much here.

I'll go check out the Digi-log. One BIG question, though--does it replicate the spring "slap-back" that happens when you palm-mute the strings and hit 'em hard? (Best example that comes to mind is in John Mayer's Slow Dancing in a Burning Room.) If it doesn't do that, then I have to stick with spring tanks.


And yeah, you're right--reverb with SE doesn't sound all that great, but again, I've got those crazy constraints on size and loudness... I'd love to keep my Princeton Reverb clone (sounds amazing), but it doesn't fit into our modular shelving like a Champ does.

Thanks for all the help, BiBi!

Oh, to anybody who happens across this thread, now that the circuit is beginning to deviate too much from what is likely to be the Eric Barbour reverb, I'll start posting newer schematics separately from Rev. 1.0a.

[dchang0]

To anyone following us down the path of using the triode-strapped pentode of the 6BM8, here is a link to hand-collected curves of that mode of operation:

http://www.tubes.mynetcologne.de/roe..._as_triode.pdf

Okay guys, I'm just thinking out loud here.

From The Valve Wizard:

"Accutronics quote the saturation current of their cores as 3.5Amps per turn (rms),...

The voltage required to achieve the Nominal Drive Current can be found from Ohm's law:
V = IZ
Where Z is the impedance of the coil, which is normally quoted at 1kHz.

For example, if a type A tank is rated as 8 ohms at 1kHz and has 124 turns;
I(nominal) = Amps per turn / number of turns.
I(nominal) = 3.5 / 124
= 28mA"

Now, the curves on the triode-strapped pentode 6BM8 indicate that it can easily provide around 40mA to 55mA in the middle of its curves. That's almost double the 28mA needed for the 8A-model 8-ohm input reverb tank. But Rod Elliot notes "After getting a new 8 ohm tank for some experiments and to take a few measurements, it turns out that the coil can be driven somewhat harder than claimed. I was able to drive the 8 ohm coil to 250mA at 1kHz before saturation (almost 10 times the current claimed). The saturation current remains roughly the same at all frequencies from around 300Hz and up, and at 1kHz the voltage was measured at 2V RMS."

So while the 55mA output of the triode-strapped pentode 6BM8 might be more than the rated saturation current, it's not enough to truly saturate it.

Looks like I'll HAVE to buy the 8-ohm tank, as the 8B-model 160-ohm's nominal drive current is a mere 6.5mA, which can easily be driven to 10x = 65mA by the 6BM8.
-----

Looking at the chart, It would probably be better off choosing a load curve closer to 40mA than 55mA. I selected 44mA to correspond with a 6.2K plate resistor.

-----

Again, looking at the chart, I selected a middle-of-the-road bias for the highest headroom: -14V. That's about 20.5mA quiescent current, just under the 28mA needed for the 8A tank.

-14V / 20.5mA = 682.9 ohm ~= 680 ohm for the cathode bias resistor.

Power rating?

P = RI^2 = 680 ohm * (0.0205A)^2 = 0.285W < 0.5W. Let's go with 1W to be safe.

This does raise the question: why did Eric Barbour use a 10W (750 ohm) resistor? Is my 1W rating going to be too low?

-----

Going back to power rating for the plate resistor:

P = RI^2 = 6.2K * (20.5mA)^2 = 2.605W. Crap. Gotta step up to 5W wirewound, and because of this, 6.2K is out. 6.8K is available instead, lowering the load curve to 40.4mA.

Gotta redo the calculations on the new load line. Quiescent current is now 20mA at -14V bias.

Cathode bias resistor = -14V / 20mA = 700 ohm =~ 680 ohm. No change, then.