18Watt.com

I was thinking of building something with ZP's cascade/parallel switching single channel amp. Ofcourse I want reverb and tremolo on board.
I made a schematic of what I think is the circuit for just that.
I borrowed Josh's reverb design and took the IRF840 tremolo from the imperial... both sublime designs I would like to adapt on this build.

Can anybody take a look at this? Power supply from the classic 18W in mind...

The only part of this I can really comment on is the reverb. A 12AX7 doesn't have nearly enough current capability to drive the tank to its full potential. That doesn't mean it won't be usable, only that it will be way under the drive levels used by Fender and others.

Jack

Ok. Will try with a 12at7 and find out the difference..
Thanks Jack

Bieworm wrote: ↑

Wed 01/20/21 3:08 am

Ok. Will try with a 12at7 and find out the difference..
Thanks Jack

I spent a lot of time last week digging around the various reverb circuits, studying the published signal levels and simulating the commercial designs in SPICE. I concluded that this function - driving and recovering from a tank - is problematic for anyone doing it with one dual-triode tube.

On the drive side, a 12AT7 generally has sufficient gain, but both sections need to be paralleled to create the current capacity needed to fully exploit the capabilities of the tank. I have to assume that for a given amount of reverb in the final mix, the tank is less susceptible to external noise and undesired mechanical effects when driven closer to its specified limits. It's just the signal-to-noise ratio, like all electronics. At the other end of the tank, recovery requires a lot of gain because the signal is so small. Most all-tube circuits use a 12AX7.

I don't know of any dual triode that meets the requirements of both drive and recovery. The 12AT7 comes close, but it doesn't match the capability that Fender designed into their amps (they typically use three sections of two tubes). There are ways around this issue, however, and I've decided to use this tube type as a starting point. My biggest concern is knowing it won't provide full drive to the tank.

Someone earlier mentioned the 6U8 triode/pentode, and I think that might be one of the possible solutions. The triode exhibits somewhat lower mu than the 'AT7 (40 VS 60), but it has the necessary drive capability. The pentode section can of course be "programmed" for a wide range of gains, depending on the anode resistor, and it looks like a good choice for the recovery side. The 6GH8 is another possibility for this, with a triode mu of 46 and anode resistance low enough to drive the tank properly.

The good news is that there are a number of options that are compatible with the 9-pin socket. Building with a 12AT7 and optimizing later should be relatively painless.

Jack

TriodeLuvr wrote: ↑

Wed 01/20/21 6:50 pm

I spent a lot of time last week digging around the various reverb circuits, studying the published signal levels and simulating the commercial designs in SPICE. I concluded that this function - driving and recovering from a tank - is problematic for anyone doing it with one dual-triode tube.

The only problem here is getting so caught up in theory, that you're forgetting about the reality of practice. Maybe in your quest you'll design the most efficient use of the tank ever made, but... There are thousands of people using the single valve reverb method using a single 12AX7. My improved version of that is what I use in my builds, and it's what Bieworm has in one of his. My point being that theories and spice models don't make those reverbs sounds any less delicious.

I'm not saying not to try your ideas. I'd love to see/hear you come up with a cool alternative. And then A/B it with the single valve method.

I tried it out yesterday and can't hear any obvious difference...

Bieworm wrote: ↑

Thu 01/21/21 3:48 am

I tried it out yesterday and can't hear any obvious difference...

The outcome of changing the tubes will depend in part on the signal level available to the grid of the driver. It might not be enough to "stress" the 12AX7, or on the other hand, maybe the distortion that results from insufficient tube current is being masked by the reverb effect. I think there must be a reason Fender used two 'AT7s in parallel for so many of their amps, but it's possible they overestimated the necessity of driving the tank harder. Did you change biasing on the two sections when you changed the tubes? In any event, if the effect with both tube types is good, that's great. This approach is a lot easier than screwing around with multiple tubes or oddball types. I'm hoping it works for me, too!

Jack

TriodeLuvr wrote: ↑

Thu 01/21/21 4:58 am

Bieworm wrote: ↑

Thu 01/21/21 3:48 am

I tried it out yesterday and can't hear any obvious difference...

The outcome of changing the tubes will depend in part on the signal level available to the grid of the driver. It might not be enough to "stress" the 12AX7, or on the other hand, maybe the distortion that results from insufficient tube current is being masked by the reverb effect. I think there must be a reason Fender used two 'AT7s in parallel for so many of their amps, but it's possible they overestimated the necessity of driving the tank harder. Did you change biasing on the two sections when you changed the tubes? In any event, if the effect with both tube types is good, that's great. This approach is a lot easier than screwing around with multiple tubes or oddball types. I'm hoping it works for me, too!

Jack

No biasing done. Just a quick swap. They're cathode biased anyway..

Bieworm wrote: ↑

Thu 01/21/21 7:56 am

No biasing done. Just a quick swap. They're cathode biased anyway..

Being self-biased does help the swap. With the values you've shown, the driver section will idle at about 1.7 mA with a 12AX7, or 3.6 mA with a 12AT7. (This assumes 275V B+ and 1.2K DCR at the transformer primary.) That's a pretty significant difference in terms of delivering power to the transformer.

On the recovery side, the most significant consideration is voltage gain. The 12AX7 will exhibit a gain of about 55; the 12AT7 is about 38.

Again, whether any of this is audible depends on a number of additional factors including signal levels to the reverb driver and the gain structure of the amplifier overall. It would be interesting to optimize each side of the tank and do an AB test like Josh suggested.

Jack

Maybe try a 6C10. TAD sells NOS RCAs but they’re pricey. If you want to go with 3 triodes.

crgfrench wrote: ↑

Thu 01/21/21 8:58 pm

Maybe try a 6C10. TAD sells NOS RCAs but they’re pricey. If you want to go with 3 triodes.

I just looked those up on eBay. Holy Moley they're expensive!! I remember when you couldn't give those things away!