6BM8 capacitor driven reverb

[walkman]

Hi all,

Here is a 6BM8 reverb build sourced from threads from this forum.

https://el34world.com/Forum/index.php?topic=22730.msg242857#msg242857
https://el34world.com/Forum/index.php?action=dlattach;topic=7957.0;attach=56876;image

Still messing with the mix resistor, triode stage cathode resistor, and grid leak resistor to get the right balance.
Overall sounds pretty good and give more of a depth / room presence than anything else.

The current build schematic is slightly different from the photo.

[tubenit]

That's very cool!  When you finalize the schematic, please share it and I'll put in into ARCHIVES under one tube reverbs.  Grateful for you sharing this!


With respect, Tubenit

[walkman]

Thanks for the feedback. It was a tight build!

Here it is in the back of the combo that it was built to pair with.

[pdf64]

…
Still messing with the mix resistor, ...

Well done!
To me it looks like a wet only effect, input on the ring and output on the tip; so I’m puzzled by the mix resistor reference?

As an external fx unit, the high output impedance will make the tonal balance affected by cable capacitance.

[tubeswell]

Still messing with the mix resistor, triode stage cathode resistor, and grid leak resistor to get the right balance.

When you say ‘messing with mix resistor’, is there a problem with the wet effect Level? How have you got it inserted into the main amp? Passive series FX loop? (Or active?, parallel?) How far along the main amp signal chain is the loop?  (Do you have a schematic for that?)

[PRR]

....schematic...

[walkman]

Hi All,

Here is the updated schematic with correction for the heater wiring and pentode section cathode (Thanks PRR!)

The send from the amp is after the second half (V2) of a single 12ax7 (Like Timbos 10w Stinger linked in the first post).
Reverb signal is then ‘mixed’ with the dry signal. There is some interaction between the reverb volume and the dry signal when when the reverb volume is at maximum, though it is slight and does not reduce the overall output volume significantly.

Would be good to get a greater wet signal with out reducing the dry signal level with a larger ‘mixing’ in resistor in the amp 235k (470k//470k).

Though the current setup provides a depth and body to the amp and noticeable reverb decay, rather than any sort of surf or cavernous effect.

* schematic updated - triode grid leak 1M as can be seen in photos

[tubeswell]

Still messing with the mix resistor, triode stage cathode resistor, and grid leak resistor to get the right balance.

When you say ‘messing with mix resistor’, is there a problem with the wet effect Level? How have you got it inserted into the main amp? Passive series FX loop? (Or active?, parallel?) How far along the main amp signal chain is the loop?  (Do you have a schematic for that?)

[walkman]

Still messing with the mix resistor, triode stage cathode resistor, and grid leak resistor to get the right balance.

When you say ‘messing with mix resistor’, is there a problem with the wet effect Level? How have you got it inserted into the main amp? Passive series FX loop? (Or active?, parallel?) How far along the main amp signal chain is the loop?  (Do you have a schematic for that?)

Hi Tubeswell

Yes the wet level of the reverb effect is not what I had expected. It would be good to get a greater depth of effect, as with the current configuration the reverb volume must be most of the way up for any wet effect to be heard.

Is that a problem? … no, the outcome is useful and provides a noticeable body and depth to the sound.

Does the circuit function correctly I believe so, though it may not be configured correctly to provide enough drive to the reverb tank?  Or the recovery stage may not provide enough gain with the location or method it is inserted into the amp that it is currently paired with.

Please me know how I could improve the design / implementation.

The effect is passive and parallel. The preamp is a single 12ax7 with tone and volume between the sections.

Both the dry and wet (reverb send) stem from the second half of that 12ax7 after the coupling capacitor.

Here is the schematic for the the amplifier it is currently paired with.

[PRR]

> Both the dry and wet (reverb send) stem from the second half of that 12ax7

Perhaps more to a point: they are INside the NFB loop. The nature of the beast is to try to cancel "errors" inside the loop. It can't know that you want the reverb, it will try to cancel it. Because reverb is practically incoherent, it may not cancel well. I'd still try without the NFB. (That switch may have a setting but I'm not sure.)

[walkman]

> Both the dry and wet (reverb send) stem from the second half of that 12ax7

Perhaps more to a point: they are INside the NFB loop. The nature of the beast is to try to cancel "errors" inside the loop. It can't know that you want the reverb, it will try to cancel it. Because reverb is practically incoherent, it may not cancel well. I'd still try without the NFB. (That switch may have a setting but I'm not sure.)

The amount of NFB ‘switchable’ from virtually nothing to enough to drop output by 3db at a guess, and was part of the design for the amp. The reverb was a recent / late addition. I can disconnect the NFB and see if it makes any difference.

Reverb is still present if the NFB is switched in but to a lesser degree.

* error corrected at NFB / boost switch, schematic in my previous post updated here

[pdf64]

The amount of NFB ‘switchable’ from virtually nothing to enough to drop output by 3db at a guess, …

A NFB loop requires gain within the loop for it to take effect. Hence the effect of NFB with a gain control inside the loop will depend on the gain control setting. So with the master volume set high, the NFB switch should have a bigger effect than with the master set low.

[walkman]

I mostly play the amp without the NFB switched in.

The way the switch is configured it allows bass boost normal or low / flat tone. Works quite well with the MV 2/3 way up.

[tubeswell]

The reverb effect insertion point is at the output end of a voltage divider that’s following a 2nd preamp stage, so:

1) the dry level signal that’s hitting the output tube is attenuated quite a bit compared a normal champ type amp anyway (thus weakening your amp’s dry signal mojo), and

2) the wet signal from the reverb recovery stage is quite a bit weaker than a signal from a normal input stage, due to the very weak input signal from the reverb pan’s output transducer (compared to the signal from a typical guitar pup - even a single coil pup). This is competing with a much stronger dry signal from the amp’s 2 preamp stages (even though you have the dry signal knocked down a bit with the voltage divider across the FX insertion point).

[walkman]

Thanks for the explanation Tubeswell.

Is there much that can be done to improve the reverb level at the current insertion point?

Or is it preferable to change the insertion point?

[sluckey]

I don't see any options that don't require an additional gain stage. Boosting the wet signal requires another gain stage in the recovery circuit. Lowering the dry signal using a big mix resistor will produce a bigger wet/dry ratio but also requires another gain stage to boost the overall signal back to a proper level to drive the EL84.

[walkman]

Perhaps replacing the EL84 with a 6GW8 or 6BM8 would be the trick. Though that may be a different amp build.

As is the amp sounds good with the reverb at max even though the level is low.

The drop in level compared with the that full dry level ( when the reverb is not inserted) is noticeable, however not detrimental to the tone of the amp.

With the lead to the reverb removed the amp gets the full dry signal.

A different design for the reverb for this amp with an extra recovery stage as suggested would be the better option.

[PRR]

Op-amp chip.

Drive the tank with a 47r resistor from the speaker output and bring it back at an early stage.

Plagiarize!!! There are hundreds of reverb-amp schematics. How many work with this few tubes? Unless you are smarter/cheaper than Fender or Matchless or DanElectro, you are unlikely to beat their best.

[walkman]

Op-amp chip.

Drive the tank with a 47r resistor from the speaker output and bring it back at an early stage.

Plagiarize!!! There are hundreds of reverb-amp schematics. How many work with this few tubes? Unless you are smarter/cheaper than Fender or Matchless or DanElectro, you are unlikely to beat their best.

Not trying to out gun the big guns like fender and alike, simply discovering more about the wonderfull world of tubes ;-)
This is my first voyage into reverbs, so happy to try different approaches to get different results.

2) the wet signal from the reverb recovery stage is quite a bit weaker than a signal from a normal input stage, due to the very weak input signal from the reverb pan’s output transducer (compared to the signal from a typical guitar pup - even a single coil pup).

It occurred to me this morning that if the triode plate resistor was increased to 270k and the cathode resistor lowered to 330R there may be a little more drive to the tank and a little more recovery gain. I wonder if the grid leak can be increased on the triode section to 2M or 5M with out drawing excessive grid current.

Can grid current be prevented by referencing the cathode (or grid) to the plate ( current feedback ?)

[tubeswell]

Can grid current be prevented by referencing the ...(grid) to the plate ( current feedback ?)

It's called local negative feedback and it will make the stage smoother and more HiFi, but quieter...http://www.valvewizard.co.uk/localfeedback.html

[walkman]

The transformer used in this reverb came from an old and defunct tube record player.

The valve lineup was a 6GW8 and 6V4. This leaves 600ma of 6.3v that could perhaps be rectified or voltage doubled to power a FET or MOSFET or something else.

There is enough room to mount a little tag strip or veroboard on the side behind the input / output jack.

Haven’t yet searched for ideas so any ideas would be great.

Would a FET design like runoffgroove fetzer with a fixed gain as the first reverb recovery stage work?

Or better to consider a fixed gain stage after the reverb volume?

Not that there is much space left to add a 12ax7 but I’m tempted to add one to the top / base plate.

With the 6BM8 running triode strapped there is around 10-12ma of B+ that could be free for use. Would be tricky but might squeeze in. What do you think?

[pdf64]

…
With the 6BM8 running triode strapped …

Triode output stages typically have lower gain and power output than pentode. What’s the rationale for the triode arrangement?

A LND150 MOSFET can be used for an additional, ‘12AX7 like’ gain stage, they’re common in fx loops.

[walkman]

Triode output stages typically have lower gain and power output than pentode. What’s the rationale for the triode arrangement?

The build was sourced from the two threads linked in the first post. It was my understanding that these were proven designs. The circuit does work however is not optimal in the application for the amp I have paired it with