Questions regarding the 6G15 Reverb Unit circuit / design.

[Baguette]

Hello,

I'm considering building a 6G15 Reverb box and the schemo makes me wonder about some things.

1/ I'll build mine using a bridge rectifier. Could I replace the input filter choke (that filters the whole B+) by a high power resistor? What value would give me the same filtering effect a the 4H choke?

2/ Seems like the 6K6 screen is connected directly to B+... how about that? Why isn't it connected the usual way (e.g. Champ style: plate to the OT, then RC filter, then screen)? Does this configuration change the sound?

3/ The input signal goes to a 12AT7 which is wired like this:
> 12AT7 triode w/ split plate load > Dwell > 12AT7 triode > 6K6
Why the heck Leo did it like this? Seems pretty weird to use two cascaded stages of a low gain tube and knock the signal down even more with the split plate load... why didn't Leo use a single 12AX7 stage to feed the 6K6? (which btw is the way it's been done on the Fender VibroKing).

Thanks for your help!
 

[Baguette]

Any insight?

[tubeswell]

I made a 6G15 clone a couple of years ago and used a PT with 240-0-240 @100mA and a FW SS rectifier. But you need that little fender choke http://triodestore.com/22707.html, because you want to filter out as much residual ripple as you can in this circuit (remember the RT is an SE output transformer, so like any SE amp, it'll hum more if you don't use a choke). The unit sits in front of another amp, therefore getting a really low noise:signal ratio at the unit's output is vital (without wanting to sound too hysterical about it).

Similarly, I found that using 1kV .01uF caps in parallel with each diode in the rectifier quietened down the diode-switching noise (and I conducted my own controlled experiment removing and replacing the caps just to make sure they reduced the switching buzz, and they did. I used 2 x 1N4007 in series on each side of the HT winding, each diode in parallel with a cap (i.e.; 4 x diodes and 4x caps in total). Another method could be to use fast-switching diodes.  There is some controversy from skeptics about the benefits of snubber caps on diodes in this regard, but I recommend them if you are using 1N4007s, because it worked for me. Its all about the SN ratio.

The other thing that's important is grounding. I used the system in R.G. Keen's (attached) article and the thing is pretty damn quiet. JM2CW

And another 'nother thaing :-) I'd use an RT with a 5k Pr Z. I tried both 5k and 8k in mine and found the 5k a lot cleaner and gives better tonal results. Using 8k is overkill and you just end up with mega distortion

And one final 'nother other thing. Use properly shielded transformers and choke. Everything sits in close proximity to the pan transducers in those confined spaces, so you want to reduce unwanted EMF coupling as much as possible

[sluckey]

1/ I'll build mine using a bridge rectifier. Could I replace the input filter choke (that filters the whole B+) by a high power resistor? What value would give me the same filtering effect a the 4H choke?

You can do that, but no single resistor will give you the same filtering effect as the choke. You can get close by adding another RC node. I'd just use the choke.

2/ Seems like the 6K6 screen is connected directly to B+... how about that? Why isn't it connected the usual way (e.g. Champ style: plate to the OT, then RC filter, then screen)? Does this configuration change the sound?

Who knows. Leo thought it worked pretty good.

3/ The input signal goes to a 12AT7 which is wired like this:
> 12AT7 triode w/ split plate load > Dwell > 12AT7 triode > 6K6
Why the heck Leo did it like this? Seems pretty weird to use two cascaded stages of a low gain tube and knock the signal down even more with the split plate load... why didn't Leo use a single 12AX7 stage to feed the 6K6? (which btw is the way it's been done on the Fender VibroKing).

If you look closely you'll see that approx 90% of the signal passes thru that split plate load. Using a single 12Ax7 stage will waste the other triode. Who can say why Leo chose this design.

[Baguette]

Thanks for your replies, that's much appreciated.

I went with the 125C3A (Deluxe type) choke + bridge rectifier.
There's zero hum, quite impressive for a single ended amp.

As for the power tube wiring, I think it's the first time I see a pentode with the screen at a higher potential than the plate. Just wondering about that... but it sure does the job.

I totally changed the wet side topology, going from 2 12AT7 triodes in series to 2 12AX7 triodes in //.
That's closer to the VibroKing approach (except there's only 1 12AX7 triode in the VK).
I think it sounds way better this way. Like, you can crank the Dwell and have big full wet reverb, but still sounding totally CLEAN. The standard Reverb Units I've tried have some cold raspiness when the Dwell is up, kinda like the second 12AT7 stage is clipping (and I don't particularly like the tone of a clipping 12AT7). Anyways, I'm happy like that.