Single tube reverb revisited- long boring post

[imaradiostar]

I have been messing with some single tube reverb ideas and I thought I'd share in hopes that someone else out there can use it. I grafted verb into a root 666ish amp that I built and I think it would be easy enough to adapt it to nearly any amp if you can spare .45a of fil current and the space for one more 9 pin tube.

I was digging through my piles of junk and I found a few tubes that contain a triode and a sharp cutoff pentode in one bottle. I grabbed a 6an8a (for no particular reason) and built a simple circuit driving an old hi-z input reverb tank I had laying around from some old solid state amp. I have a bunch of old tanks like this around so I've been looking for a way to utilize them.

The triode side of the 6AN8 as a lot like a 12au7 and is rated at about 3 watts so I figured it's similar to the 6SN7's that ampeg used to drive reverb. I used a 330 ohm cathode resistor and a 8.2k plate resistor- splitting the difference between values I found in different ampeg circuits (2.2k to 10k on ampegs). I tapped the screen supply and this left the screen voltage around 380 volts. I drove the tank with a .82uf 630 volt cap because it was what I had available. I hadn't gotten around to adding a .022uf to ground like an ampeg but I'm going to try it when time allows. I drove the triode directly from the .022uf cap at the plate of the input stage ef86, passing through a .001uf cap and a 1m grid resistor. If I hadn't been in such a hurry this would have been a 1m pot.

I setup the pentode side using the same values as the root 666 input stage- I figured it's less microphonic than a matchless with these values so why not stick with them? I used 220 ohms from ground to grid one, attached to the verb return jack. I used a .003uf cap driving an audio taper 500k pot with the wiper tied to the other side of the phase inverter a la Matchless. I didn't bother with pretty summing resistors or tailoring frequency response because this was only a proof of concept.

So...how did it work? A little too well it seems! The triode was getting punished by the EF86- some better gain staging is in the works. I prefer to drive verb before the volume control (unlike Fender) because sometimes I still want lush full reverb with the volume down. This makes a dwell control almost a necessity. When I turned down the guitar volume so that the driver triode wasn't getting too much signal the verb was nice and clean and still had more than enough dwell. I will add that 1m dwell pot shortly...

The recovery side was a similar story. There is more than enough gain available to drive the phase inverter into distortion with reverb alone. The tone of the verb still needs to be sculpted a bit (once again, I didn't even try) but the most important thing is that there is plenty of signal and it's clean enough to make it work. I'd say there is enough signal that I could easily blend it in at the other side of the PI using a resistor if I wanted to. Mission accomplished.

Here's a short list of tubes that I found that are still available (listed tubestore.com price), affordable, can drive a high impedance tank and should work well for this kind of circuit:

6U8- 2.8 watt triode, mu=40, $6.95
6AN8A- 2.8 watt triode, mu=21, $9.95
6GH8A- 2.5 watt triode, mu=46, $7.95
6KZ8- 2.5 watt triode, mu=46, $6.95

I'd bet that others exist too...

If you try it (or already have) please respond here and share- I'd be curious to see what you found.

jamie

[tubeswell]

I haven't tried it yet but I was looking at using an ECL86 or an ECL82 for a similar role, with the pentode part driving and the triode part recovering. (I believe New Sensor or EH are still making these?) There's a bunch of discussion on 1 tube 'verb over at the Dumble forum. Somebody there sent me this circuit FWIW. I've got a bunch of ECL82s lying around, so I'll give it a go in a build or two from now.

Data on ECL 82

http://www.r-type.org/exb/exb03969.htm

Data on ECL86

http://www.r-type.org/exhib/aaa0034.htm

[tubeswell]

Eh 6bm8 / Ecl82

Knew I'd seen it somewhere

http://thetubestore.com/ehx6bm8.html

[imaradiostar]

I built an ecl86 based verb into a fender style amp about ten years ago. I thought that it didn't have enough recovery gain and I had to add another gain stage to make it sound decent. I used a transformer from an old tube hi-fi and it sounded good but not great. I'm not saying don't do it- with the right setup it should sound similar to the Fender standalone verb. As Bruce suggested elsewhere I'd try an J201 as an additional gain stage. Look up the "Fetzer Valve" at runoffgroove.com and use it as the first stage after the verb tank. I thought about putting a J201 into the RCA cable to get the recovery amp as close to the tank as possible- this wouldn't be tough but I don't know if it would be that much cleaner. When using a tank with an isolated output I bet it would be excellent.

I wonder how an ECL82 or 86 would do without a driver transformer? I just like not having a transformer because I feel like a big cap is cheaper and more than sufficient for the tiny drive requirements of a tank.

jamie

[mooreamps]

I have a question. What do you guys think of this ?
Have an EF-86 drive the input to an EL-84. Then, in the plate circuit of the EL-84 drive a Valve Junior output transformer. Then off the 8 ohm tap, use that to drive a Fender style long tank reverb ?

-g

[Bruce / Mission Amps]

I have a question. What do you guys think of this ?
Have an EF-86 drive the input to an EL-84. Then, in the plate circuit of the EL-84 drive a Valve Junior output transformer. Then off the 8 ohm tap, use that to drive a Fender style long tank reverb ?
-g

Why go through all that?
Check MOJO's MOJOVERB.
http://www.mojomusicalsupply.com/ite...77&id=MOJOVERB

Basically a small 3-5 watt amp driving it's own speakers wiht 100% reverb.

[mooreamps]

Why go through all that?
Check MOJO's MOJOVERB.
http://www.mojomusicalsupply.com/ite...77&id=MOJOVERB

Basically a small 3-5 watt amp driving it's own speakers wiht 100% reverb.

That's very impressive Bruce. I'm going to have to give this some thought.

-g

[Bruce / Mission Amps]

Just for fun, while I had a BF Fender Reverb apart last night, I undid the leads from the reverb's 12AT7 socket and used a 12DW7 in that socket instead.
The OT now was driven with the low mu section of the 12DW7 and I slipped the recovery leads from the 12AX7 recovery triode over to the 12DW7's hi mu section... guess what?
It worked very well and with the reverb level pot set to midway it still made great reverb.
At that mid way level setting, as viewed with my scope and measuring actual AC voltage out of the triode's plate, it only made about 25 to 30% less reverb the the stock Fender circuit with the level set in the same place.
But... with the single 12DW7 I still had more left on the level pot and I was quite surprised to find that the difference was ...
I liked the stock reverb sound with the level control set at 2.5 and with the single 12DW7 tube I like it set at just under 4.... still had plenty of reverb left. So much that it was still unusable for anything I could think of.

If you are just counting the 12AT7, a single tube reverb is possible.
I'm 100% sure that using a J201 FET (with a gain of around 16-18dB), in place of that 12AX7 triode would have allowed me to use the hi mu section of the 12DW7 to be the final third triode which would amplify the dry mixed with the reverb from reverb the level pot... You know, the classic Fendery, 3m3 resistor with the 10pF cap across it.
I don't have all those computer skills most of you do so I drew this on my scratch pad here just now and scanned it.... I eliminated the 100nf cap and the load down section from the intensity pot in the vibrato circuit to make it easy...

[CitizenCain]

When I rebuilt my Sundown head into a hand-wired tagboard, I converted the reverb circuit to a 1 tube affair. It gives me plenty of reverb, enought to swamp out if I want to

The original design used paralleled 12AX7 driver and 1/2 12AX7 for recovery. I eliminated the paralleled driver and use 1 12AT7 for both jobs. Seems to work out well. Can anyone tell me if this could be problematic down the road?

Here's the schemo I used. It says 12AX7 in the legend, but I used a 12AT7 because the voltages came out better.

http://www.visi.com/~sstolle/sundown/sd1000h-mod.GIF

[imaradiostar]

I forgot about this thread- I was glad to see there are a few more replies.

Bruce, I'm glad to hear that someone actually tried that circuit- I've been wanting to build it for a while. The thing you posted is exactly what I've wanted to build into a standard blackface fender with LTPI. The J201 is a pretty neat little device...

I've also got a board laid out to build a trem-free princeton reverb using a 12dw7 in place of the 12at7 and one of the halves of the 12ax7. I'm hoping to have a pair of amps built this way in the next month or two.

I've been wanting to try something I saw on the internet recently. Someone out there posted about using a reverb tank with a floating output and connecting the tank output ground directly to the source of a jfet. I wonder if this would work on a tube gain stage and result in quieter reverb.

I haven't had much time to mess with my single tube verb circuit but the real problem I've had thus far is that it's noisy and terrible! I'd imagine the improvised layout and terrible grounding scheme had a lot to do with it. I'm also not sure how to do the blending- since it doesn't use two gain stages like the Fender circuit then it's a little more confusing how to blend the verb back in...in a way that sounds decent. I hate to admit it but that huge resistor and tiny cap sound great in a Fender. I'm going to try some different ideas when I get a chance.

jamie

[CitizenCain]

Bruce, that 12DW7 idea really intrigues me. I'm thinking I might like to try that in my one-tube Sundown circuit. Would you bias the low mu side any differently than it's shown in my schematic?

I also see that most Fendery reverb circuits have a 220k resistor to ground right after the reverb tank. I suppose that's the grid resistor for the recovery triode. My schematic doesn't have one, any problems like that? That's how the original circuit was desinged by Kager in the Sundown. Also no mixer resistor, is that because it's a one-channel amp?

[Bruce / Mission Amps]

Bruce, that 12DW7 idea really intrigues me. I'm thinking I might like to try that in my one-tube Sundown circuit. Would you bias the low mu side any differently than it's shown in my schematic?

I also see that most Fendery reverb circuits have a 220k resistor to ground right after the reverb tank. I suppose that's the grid resistor for the recovery triode. My schematic doesn't have one, any problems like that? That's how the original circuit was desinged by Kager in the Sundown. Also no mixer resistor, is that because it's a one-channel amp?

The 820 ohm resistor is just what I'd start with, you might find 680, or 470 might sound better.
The idea is to get some current flowing in the reverb OT.
If you use a J201, I'd use anything from 220K to 1M on the recovery side
and at least a gate stopper, 100 ohm resistor in series with the signal to the gate lead. These J201s can be unstable and oscillate, amplifying crap right up into the radio freqs and you want to snub that possibility.
The DC resistance of the coil on that side of the tank will be in parallel with any resistor you use.
The 220K resistor actually swamps across the coil with respect to the gate or grid following it.
Some even think that the original Fender circuit, with a 220K resistor, parallel with a .002uF cap and the coil of the reverb unit, forms a small tank circuit.
A tank circuit is a coil and a capacitor in parallel which at some frequency forms a high impedance. The AC currents in the tank circuit rotate, like water sloshing around in a, well a fish tank.... and it might add to the reverb sound.
The 220K resistor would then swamp that impedance by flattening it's "Q" and broaden the frequency response.
True or not true, it sounds good...
I tend to think it is a high frequency snubber.

You might find that using a low voltage .1uF capacitor from the recovery side of the tank to a gate load resistor of 220K to 1M might be better then connecting right to it.

[Bruce / Mission Amps]

Just for fun, while I had a BF Fender Reverb apart last night, I undid the leads from the reverb's 12AT7 socket and used a 12DW7 in that socket instead.
The OT now was driven with the low mu section of the 12DW7 and I slipped the recovery leads from the 12AX7 recovery triode over to the 12DW7's hi mu section... guess what?
It worked very well and with the reverb level pot set to midway it still made great reverb.
At that mid way level setting, as viewed with my scope and measuring actual AC voltage out of the triode's plate, it only made about 25 to 30% less reverb the the stock Fender circuit with the level set in the same place.
But... with the single 12DW7 I still had more left on the level pot and I was quite surprised to find that the difference was ...
I liked the stock reverb sound with the level control set at 2.5 and with the single 12DW7 tube I like it set at just under 4.... still had plenty of reverb left. So much that it was still unusable for anything I could think of.

If you are just counting the 12AT7, a single tube reverb is possible.
I'm 100% sure that using a J201 FET (with a gain of around 16-18dB), in place of that 12AX7 triode would have allowed me to use the hi mu section of the 12DW7 to be the final third triode which would amplify the dry mixed with the reverb from reverb the level pot... You know, the classic Fendery, 3m3 resistor with the 10pF cap across it.
I don't have all those computer skills most of you do so I drew this on my scratch pad here just now and scanned it.... I eliminated the 100nf cap and the load down section from the intensity pot in the vibrato circuit to make it easy...

The reason I wanted to mock that 12DW7 up in real time was I was thinking of using it in that little 10 watt 6BM8 amp project I've been talking about on another thread. Search 6BM8
With reverb, this little amp would now have a total of 4 tubes.
Two of which are the actual power tubes in push pull class AB... the 12Dw7 would be the single reverb tube and reverb would mix in back at the phase inverter for a slightly below average reverb signal.
On paper it would work great.

[CitizenCain]

Well, I'm thinking of trying it without the J201, just subbing the 12DW7 in place of the two existing triodes I'm using.

I might try experimenting with that swamp circuit too, sound like a neat concept.

[Bruce / Mission Amps]

Well, I'm thinking of trying it without the J201, just subbing the 12DW7 in place of the two existing triodes I'm using.

If using an LTP type, that's why I mentioned inserting reverb from the level pot back into the other side of the phase inverter, ... that triode still has gain.