Help to "Fenderize" reverb on GA-1 RVT

Discussion in 'Shock Brother's DIY Amps' started by NSB_Chris, Jul 4, 2019.

  1. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    54
    Posts:
    133
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    I would really benefit from help from those that have knowledge and experience with reverb circuits. I have a few threads on this amp project. Latest: http://www.tdpri.com/threads/gibson-maestro-ga-1-rvt-build-input-needed.953848/#post-9123825

    I have done a lot of research on this and anybody that has had issues with the vintage reverb circuit has come up very short. The driver transformer and the reverb tank are mysteries. All have difficulty identifying what the originals are and finding replacements. Since I am building a new amp and not altering a vintage amp, I can do whatever is necessary. I am considering shoehorning in a fender reverb circuit but I don't have knowledge and experience with them. Hoping others that do will weigh in.

    Since this amp has dual pentode-triode tubes, I am going to try to use the existing reverb driver and recovery tube sections. I think the reverb driver and recovery triodes are essentially low noise 12AX7.

    I am attaching the original circuit (redrawn) and my redesign options 1 & 2.

    Option 1 uses basically the same strategy as the original and just substitutes a fender driver transformer and tank. The tank seems reasonably close to the original (short 2 spring). The fender reverb circuits that I referenced used an input coupling cap and a grid leak resistor to the driver tube. Not sure this is necessary with the original circuit. One big problem with this option is I can't seem to find the 2M-A pot for the reverb level.

    For option 2, I basically tried to use the Fender reverb circuit except for using the original driver and recovery tube setup, which I am hoping will work. This one required a pretty drastic change of the volume arrangement between the first pentode stage and the second triode gain stage. Seems this should work just fine and I can source all of the components!

    GA-1 RVT V01.jpg

    Option 1:
    GA-1 RVT V02A.jpg

    Option 2:
    GA-1 RVT V02B.jpg
     
    Last edited: Jul 5, 2019
  2. tubelectron

    tubelectron Tele-Meister

    Posts:
    377
    Joined:
    Nov 25, 2010
    Location:
    France
    @NSB_Chris,

    Personally, I'd choose the Version 3.

    That said, it's under reserve of a satisfactory test that I strongly recommend to avoid further bad surprises...

    I would decrease the 3M3 to 1M5 and increase the // cap from 10p to 22 or 27pF.

    The problem is that the reverb circuit will produce a drop in signal level that you may recover in the following stages.

    Another issue is that the reverb circuit needs enough amount of drive to sound correctly at its driver input.

    That 's why I recommend testing...

    Here's a reverb circuit that works well on one of my DIY amps, and may work for your design :

    [​IMG]

    [​IMG]

    This reverb circuit is preceeded by the classic Fender 2 stages tone and volume circuit as found on a Twin Reverb.

    This reverb circuit is followed by a triode gain stage tied to a cathodyne splitter à la Princeton Reverb, then a push of 6L6WGB for circa 25W RMS output.

    Your pentode input stage without tone control may be able to provide the expected front-end gain, but it would be better to check.

    The very small Archer transformer is a 230V/6.3V 0A3, but the usual 022921 Fender reverb Xfo will sure do.

    The reverb that I use is a Belton coming from a VOX or a Blues Junior, I don't remember, sorry. It's a short one, due to amp compactness - shown below side to side with a Champion 600RI :

    [​IMG]

    The reverb unit is installed in a loose DIY damping bag :

    [​IMG]

    Good Luck ! ;)

    -tbln
     
    Last edited: Jul 4, 2019
    King Fan, NSB_Chris and Paul-T like this.
  3. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    54
    Posts:
    133
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    Thank you! Some additional questions if you don't mind:

    Is the capacitor between the driver tube plate and the transformer necessary? What capacitance value are you showing?

    It looks like you are dropping the B+2 voltage before the transformer. Based on the original schematic and voltages posted by someone on their GA-1 RVT, the B+2 should be between 230V and 270V, which is probably already a little low.

    Do you know why there is a 2nF cap at the recovery tube grid leak? Is it needed to bleed higher frequencies to ground after the tank? I have not seen this on other circuits.

    Why is there a 300k resistor to ground at the exit of the circuit? If that is a grid leak for the gain stage that follows, doesn't the reverb pot already serve as a grid leak?
     
  4. tubeswell

    tubeswell Friend of Leo's

    Posts:
    2,365
    Joined:
    Jul 1, 2008
    Location:
    NZ
    ‘Option 2’ (the 3rd schematic) is more like a traditional Fender reverb (than the others, but still isn’t). The main issue with getting a decent amount of splash is getting a good ‘balance’ between dry and wet signals. If there is too much gain in the dry signal path, and not enough in the wet signal, you won’t get much/any reverb effect.
     
    NSB_Chris likes this.
  5. Snfoilhat

    Snfoilhat Tele-Afflicted

    Age:
    38
    Posts:
    1,085
    Joined:
    Apr 8, 2016
    Location:
    Oakland, CA
    1. https://robrobinette.com/AB763_Modifications.htm#Reduce_Reverb_Noise

    There is also some info I remember seeing on a surf guitar website when I was building my stand-alone reverb that talked in depth about tweaking this cap not for noise reduction but as a EQ control for the spring reverb,

    2. It may be important to remember that the wet 'channel' is not separate -- the resistor you are looking at can have just as much effect (or more) on the dry channel as it does on the wet. The dry channel signal passes through one or more large resistances -- in the Gibson amp it's huge pots -- in a Fender it's the 3.3M with the tiny cap across it -- in tubelectron's schematic it's the 1.5M

    After passing though these large resistances, any path to ground acts as a voltage divider. That's what the various resistors to ground are for. While the Reverb Level pot also goes to ground, using this pot alone in a design would probably be a poor compromise. It has to do at least 4 jobs! That pot is a major part of the load the reverb recovery gain stage is driving, it has to act like a potentiometer giving a good sweep of reverb level, it may be the next gain stage's grid leak as you point out, and it forms part of the voltage divider that attentuates the dry signal. You can see in all the designs that there are a number of ways to put this all together that work more or less well, but none of them is super simple. The designers must have found that having multiple paths to ground did a better job of balancing those different jobs for better all around performance at the cost of a higher part-count and a more complex build.


    The wet channel is not that hard to make work correctly. I suggest reading through all of Vavlewizard's pages on reverb. But if you want your amp to sound like the orginal Gibson, then I would work on preserving as well as you can how the dry signal would be processed, the basic gain structure and EQ that the parts that appear to be 'wet channel' are contributing to, even if the reverb is generated, amplified, and mixed in differently, as in your new designs.

    tl;dr the paths to ground in the wet help to voice the dry. Put your hand over most of the wet channel but get the dry to look and work right by duplicating those paths to ground as closely as you can while also making some reverb happen.
     
    King Fan and NSB_Chris like this.
  6. tubelectron

    tubelectron Tele-Meister

    Posts:
    377
    Joined:
    Nov 25, 2010
    Location:
    France
    You welcome ! ;)

    No. It's optional. It's often found on SF Fender amps. It tames the treble frequency, giving more a large hall kind of sound rather than a spring sound. The value is 470pF.

    I do this on all my amps to improve hum filtering on plate voltage supply to the driver. If not enough filtered, you may have some hum entering the reverb circuit there. It's notably the case on BF, SF and some RI Fenders, where I add this little filter.

    That said, it also depends on where you pick the supply on the HV (B+) chain, so you may not need it. I must confess that I'm very picky about adverse hum and noise in amps...:confused:. When you take the voltage on the screen supply of a push-pull, that additional filter makes the difference.

    It is for the same reasons as the tube driver 470pF cap plate. Also found on SF Fenders factory-mounted. Value is 2nF. And it's optional, according to your taste and the reverb tank tone itself.

    Yes and no... Associated with the serial 470K, it acts an attenuator of the reverb circuit output, it eases spreading the reverb level adjustment. It also allows to make the grid return of the following stage at the right place in the layout, avoiding hum picking by induction in GND returns.

    That said, it's not compulsory - it depends on the design and layout of your circuit.

    Hope I'm clear ! o_O:)

    -tbln
     
    Last edited: Jul 5, 2019
    NSB_Chris likes this.
  7. printer2

    printer2 Poster Extraordinaire

    Posts:
    6,294
    Joined:
    May 24, 2010
    Location:
    Canada
    OK, this might seem right off the wall, Use option 2 up to the volume pot, feed the pot to the second stage. Take c11 off the volume pot, then through a 1M reverb pot into the reverb circuit. Now the odd part. Rather than a 2.2k resistor in the second cathode and a 2.2k resistor in the reverb recovery triode, run both cathodes into a 1.2k shared resistor. Cut the NFB resistor value in half to keep the voltage divider ratio the same. So no dumping gain using a 3.3M resistor, the reverb signal changed in relationship to the volume control. Use a on-off switch at the recovery triode grid to turn on or off the signal. I have always wanted to try it but have not had the time.
     
    NSB_Chris likes this.
  8. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    54
    Posts:
    133
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    Thanks everyone!!! I think on first try I will build with components and values close to what Tubelectron referenced. I will put the key components on separate turrets to make it easy to make adjustments. I think there is going to be a lot of adjustment necessary! Luckily my friend shipped me one of his three originals, so I have the real thing to compare to.

    IMG_3424.JPG IMG_3425.JPG IMG_3426.JPG IMG_3427.JPG
     
  9. tubelectron

    tubelectron Tele-Meister

    Posts:
    377
    Joined:
    Nov 25, 2010
    Location:
    France
    Yes, you will have certainly adjustments to do : you're going the safe way... :cool:

    A comment : I don't know if the cab of your amp will allow a long tank reverb unit to enter it, but if it's doable, preferably choose a long unit (serie 4 or 9) rather than a short one (serie 8). I often noticed that long units sound better than short ones, and it's difficult to find good sounding short ones :confused:... That said, the recent Belton/Accutronics short units seems to be better sounding.

    I also worked on a single tube reverb to installed in a Grestch G5222 / Champion 600 amplifer, made several circuits and recorded samples - I'll try to retrieve it, if you are interested...

    Designing a simple, mono-tube reverb circuit that sounds good at all settings without diminishing the performance of the amplifier is not so easy !

    -tbln
     
    Last edited: Jul 5, 2019
  10. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    54
    Posts:
    133
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    I will be making the cab myself so I am not restricted to the small tank. I will look at the dimensions and see.

    I suppose I could just decide to not use the triode side of the 6U8A and just add a 12AT7 tube for the reverb driver to honor the Fender circuit more closely. Then any adjustment would mostly be limited to the mixing section of the dry and wet. Even if I don't do that initially, I might cut an extra 9 pin socket hole in case I am forced to change later.
     
  11. tubelectron

    tubelectron Tele-Meister

    Posts:
    377
    Joined:
    Nov 25, 2010
    Location:
    France
    Why not ? Yes ! ;)

    But as you have an original, why not trying to test "on the fly" the circuit you project, in order to see if it works satisfactorily, and by the way set it up as you expect ? o_O

    But it's me, OK ? :D

    -tbln
     
  12. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    54
    Posts:
    133
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    The build has started!

    IMG_2359.JPG
     
  13. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    54
    Posts:
    133
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    I don't think there is much interest in this build because it is not a well known or popular circuit, but I will continue to post progress for fun. Got it mostly together except for a few additional resistors and capacitors that are on order and due to arrive this weekend. Good time to stop and trace every connection to make sure that it is correct to the circuit!

    IMG_2360.JPG
     
    RatBug and RollingBender like this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.