Current Driven Transfornerless Reverb Utilising 12AU7 for Drive and Recovery

[allyshake]

Hello, I am trying to build a tube guitar amp based on a vintage Vox circuit. I would like to use the circuit as is or modify it as a transformer-less constant-current reverb driver.

The vintage amp used two, 1 volt output, ACOS GR71 or two Sonotone 2T crystal phono cartridges for drive and receive transducers in its single light spring reverb.

The crystal cartridges are no longer available in the same format and in any case did not produce good reverb.

The circuit utilised a single 12AU7 for drive and recovery.

Ideally I want to use the same circuit to drive what I understand from my research would need to be a relatively highZ in and highZ out reverb tank.

However I don’t know how to do the math to work out this circuits Z Out from V3a and Z In at the grid of V3b.

I have tried to understand this article here: http://dougcircuits.com/reverbtanks.html

Would Z Out be 100K || (6500+1500) = 7407 Ohms? (6500 being the plate resistance of a 12AU7 Unit 1)

Would Z In be 47K || (7700+3300) = 8914 Ohms? (7700 being the plate resistance of a 12AU7 Unit 2)

I have also looked at this article http://www.channelroadamps.com/articles/reverb_driver/ where the author adds “we must make sure that the output impedance is high enough to qualify as "almost constant-current". It must be large relative to the impedance of the tank at the highest frequency of interest. Guitar speakers generally roll off everything above about 5 KHz, so we will use that as the maximum frequency."

Do you know of any stock tanks which would work in the circuit as is? With my calculations would a tank around Z In of 300 suit? At 5KHz this would give an impedance of 1500 Ohms. And Z Out for the tube at 7407 is close to 5X that of the tank at the highest frequency of interest.

If the circuit needs to be modified to suit a stock tank what changes would be necessary, (I would like to maintain the use of the 12AU7 tube if I can though, however I believe a 12AU7/12AX7 combined one for driver/recovery might be better)?

I am unsure of what Z Out the tank should be.

Interestingly there is no grid resister on V3b nor a bypass capacitors on the cathodes of V3a or V3b.

Note the high relative values of R15 and R17 and the equally small value of C6 and the coupling cap C9.

You can probably tell my knowledge is self taught. I was invalided out of the army after 16 years and use this to keep my brain from atrophy.

I am also modelling my project in 3D to help me trace the wiring paths from photo ref.

I have attached some photos of my 3D modelling for interest.

Any guidance you can give me would be most appreciated.

kind regards

https://www.dropbox.com/s/a18x1yymw8...matic.jpg?dl=0

https://www.dropbox.com/s/au8jn7lay7...nning.JPG?dl=0

https://www.dropbox.com/s/xl0vvol5w1...tial.jpeg?dl=0

https://www.dropbox.com/s/bivocojij7...AYOUT.jpg?dl=0

[R.G.]

There is a great discussion of how to drive reverb coils in the Floobydust (sic) section of the old National Semiconductor Audio Handbook. Very worth finding and reading.

Coils obey the law that says V=Ldi/dt. This means that the current in the coil falls linearly with frequency when fed with a low impedanc but fixed-volume signal. But the twist on the reverb spring is linear with coil current, so a voltage source which is linear with frequency has a falling high frequency content at the pickup end of the reverb.

You can get a constant current with frequency two ways. First, you can convert the signal voltage to a current, and drive the coil with constant current with rising frequency. Or you can drive it with a low impedance (i.e. voltage source) driver that has a linearly rising with frequency response, and achieve the same result as a current drive, but with familiar voltage-amplification technology. Both result in the same thing in the coil.

The National Semi approach was to use filters to make a constantly-rising voltage response, so they could use their IC power amp chips and easy filtering, especially since they introduced a bass cut to reduce muddy bass in the reverb, and a very high end falloff to reduce hiss.

Of course, National was working with inductive input tanks. There are still vintage piezo cartridges to be had in the UK and Europe, where this kind of thing was more popular. You could, with some effort get the cartridges, if not the transformers. Piezo cartridges are linear with frequency in -voltage- but can't drive much current, so they provide good response into a high impedance, but lose bass badly into low impedances. It's kind of the reverse of the inductive reverb coil. The Conqueror I modded to use a coil reverb tank had about a 10:1 voltage step up at the driver IIRC. I open-circuited the secondary and used a capacitor off the primary to provide a shunt0-fed voltage to the reverb tank, then tinkered with the recovery filtering til it sounded right.

I think you're going to have to give up a bit on using the original circuit. It is undoubtedly possible, but will take a lot of time and tinkering to get right with a different reverb tank. For ease of construction, have you considered putting in a Fender-ish reverb transformer and copying the Fender reverb setup?

[Leo_Gnardo]

There is a great discussion of how to drive reverb coils in the Floobydust (sic) section of the old National Semiconductor Audio Handbook. Very worth finding and reading.

HA! Floobydust indeed! I remember reading that around 40 years ago, somewhere I still have the NS handbook. Time for a re read. Thanks for reminding, R.G.

[ric]

I've had success using the cathode of a 12au7 to drive a reverb tank. Got the idea from a Gibson RVT Explorer.

[Leo_Gnardo]

I've had success using the cathode of a 12au7 to drive a reverb tank. Got the idea from a Gibson RVT Explorer.

That, and 60's Ampegs used 6CG7/6FQ7, drove straight off the plate thru a DC blocking cap. Plus Silvertones too but their reverb was never anything to write home about.

[allyshake]

I think you're going to have to give up a bit on using the original circuit. It is undoubtedly possible, but will take a lot of time and tinkering to get right with a different reverb tank. For ease of construction, have you considered putting in a Fender-ish reverb transformer and copying the Fender reverb setup?

Thanks R.G. Whilst the Fender transformer driven circuit is tried and true my preference is to stay true to the original if that is possible. That said, changing tube type is a distinct possibility to achieve the aim, but a 12AU7 or 12AX7 would be ideal if I can make it work with appropriate changes to resistor and cap values. In a similar vain I wouldn't go down the OpAmp path even though there are plenty of more modern reverb circuits out there that work well in practice.

I got a message from someone who said a tech used a 4F tank with the original circuit with good results. He later decreased R16 from 1.5K to 820K, which had the effect of driving the reverb tank harder. However this was the limit of info about any other changes he may have made.

The problem is that I am limited by my knowledge. Despite doing an armoured vet design course in the late 80's and having a good tech bent, I am late to the game as it were but determined to learn.

For E.g I don't know why in my circuit there are no bypass caps at R16 or R18, like there are elsewhere. Is this because the biasing is achieved without the need for them?

Similarly I don't know why there is no Grid Leak resistor on the grid of V3b. Someone suggested a 1Meg R would be appropriate there.

Thanks for your input it is appreciated.

[allyshake]

That, and 60's Ampegs used 6CG7/6FQ7, drove straight off the plate thru a DC blocking cap. Plus Silvertones too but their reverb was never anything to write home about.

Thanks Leo, is there a specific Amp amp model that I can look up this circuit? Doesn't the Silvertone use a transformer on the drive side? If not which circuit should I consult?

I found this thread http://archive.ampage.org/threads/2/..._socket-1.html which compares the similarities between 12AU7 and 6CG7 and so the Amp circuit might be a good place to start

[allyshake]

There are still vintage piezo cartridges to be had in the UK and Europe, where this kind of thing was more popular. You could, with some effort get the cartridges, if not the transformers. Piezo cartridges are linear with frequency in -voltage- but can't drive much current, so they provide good response into a high impedance, but lose bass badly into low impedances. It's kind of the reverse of the inductive reverb coil.

I have looked for alternative cartridges but without success. In any event Jennings used these in order to circumvent having to pay a licence for the US built tanks of the day. The original cartridge system never worked well and in most of the few original amps left in existence, they are inoperable. Effectively a cheap and nasty alternative which never produced good results.

[allyshake]

After some more googling... I found a schematic for an Ampeg Gemini G12 which uses both sections of a 6CG7 to drive the tank and half a 12AX7 for recovery.

This is similar in some respects to the Valvewizards article where he mentions cheap Current Source Drivers, see http://www.valvewizard.co.uk/reverbdriver.html

That would leave me with half a tube short if I used both sections of the V3 tube in my circuit for tank drive.

So could I use a 6AV11 which is effectively 3 x 12AU7's in one compactron tube? I could modify the H.T. supply like the valve wizard circuit and the additional heater current demand is only 0.3 A.

Now forgive me if I am showing complete ignorance here, as I said I am learning as I go, but have a lateral thinking brain when it comes to problem solving...this is possibly not the best thing in this circumstance, I don't know.

If this is likely to work and someone could help me draw an appropriate circuit showing the right values of caps and resistors to use that would be most helpful.

Cheers

[Leo_Gnardo]

Good research you did today, sorry I wasn't around to answer.

A tube that may get you out of your quandary, 12DW7/7247 now JJ sells their version ECC832, reasonably priced. Drive with the triode that resembles 12AU7, recover with the hi mu triode resembling 12AX7. Might have to do a little hi frequency trimming if it starts to whistle due to feedback because both drive & recovery are within the same tube and might "see" each other thru capacitance within the tube envelope.

I'd avoid that triple triode. Might have a hard time finding a socket, plus they're getting scarce & if you need a replacement in the future might be the devil to find. This is now a problem for those triples Ampeg used in the 60's. They were already obsolete then, intended to build color TV's but need for them vanished when transistor TV's became the norm. Ampeg probably bought them dirt cheap.

[allyshake]

Good research you did today, sorry I wasn't around to answer.

A tube that may get you out of your quandary, 12DW7/7247

I'd avoid that triple triode.

Thanks Leo, good advice re the triple triodes. Given the Ampeg uses both sections of the 6CG7, (which actually resembles a 12AU7 closely) and the Valvewizards similar solution of both sections of an ECC82 (12AU7) for the drive portion, might I run out of juice so to speak using half of the 12DW7/7247 tube for the drive circuit?

Perhaps the suggestion of Channel Road Amplification article cited earlier using the pentode section of a 6GH8A for drive, given its high plate resistance would be better. Valve Wizard mentions the use of "a larger triode like an ECC99 or even a triode-strapped pentode might be used on its own" since "we must use both triodes to obtain sufficient power since no OT is used".

Failing that I could use both sides of my 12AU7 for drive and add another single triode tube or half of another 12AU7. This would leave a free section for something else, although I don't know what? And that takes me quite a distance from the original I am trying to stay true to.

The search continues.

[Justin Thomas]

There are lots of Ampegs that used a 6U10 triple triode, which I believe is either a 12AT7 & 1/2 12AXY or a 12AX7 & 1/2 12AU7 or something. Point being, it works; look there for component values.

To be fair, you may have difficulty finding non-PCB-mount Compactron sockets unless you scrounge.

Justin

[Gingertube]

ECC99 is a good tube but is really "just" a bit more linear 12AU7 with 3.5W vs 2.75W plate dissipation.

So for circuits it is just an option (a good one) to use in place of a 12AU7 or 6CG7, it doesn't offer any new circuit options.

My experience with ECC99 was a simple replacement of a 12AU7 driver in a some parallel push pull triode strapped EL34 Hifi Monoblocks.
In that instance it was "subjectively" quite a bit better.

For a Git Amp Reverb is a very attractive Effect.
It is part of the provision for effects.
When designing in an effects loop you consider that everything you likely to put in that loop is going to be Solid State so my own view is that you might as well save some complexity and use Solid State for the Reverb Driver and Recovery too.


Now to shoot myself in the foot - what is in my own personal home brew Git Amp?
Kevin O'Connors (London Power) circuit using a push pull 12AT7 driver with 12AX7 Recovery. Accutronics 9 Series Tank (Medium Delay) That is seriously lovely.

Cheers,
Ian

[allyshake]

Kevin O'Connors (London Power) circuit using a push pull 12AT7 driver with 12AX7 Recovery. Accutronics 9 Series Tank (Medium Delay) That is seriously lovely.

Thanks Ian, if I can avoid using SS then going the extra mile and complexity is OK. I assume the LP circuit you mentioned in your own amp is in one of his books?

Is it something I could inject into my circuit in place of the 12AU7 V3 tube arrangement?

Is there a schematic anywhere?

Have had a quick look at LP website, does this circuit need a OT? Something I wish to avoid if possible.

regards

[g1]

What is the Vox amp model we are talking about? (https://www.dropbox.com/s/a18x1yymw8...matic.jpg?dl=0)
Can someone tell me how the dry signal path goes? I'm seeing everything going through the tank so I'm missing something.
I'm wondering if there is an extra recovery triode due to the use of phone cartridges. If so, that would free up an extra triode which would enable parallel triode drive ala Traynor YGM3.