is there a way to add a trainwreck style gain stage to a princeton reverb?

[phsyconoodler]

I want to add a gain stage to a Princeton reverb circuit and looking at the Trainwreck Express it looks like it might be the easiest.
Maybe Tubenit could draw a schematic and layout for that?
  I'm having trouble visualizing where to insert it without disrupting the reverb.

[tubenit]

Physconoodler,

Just to be clear ......... are you wanting to use a typical Princeton Reverb cathodyne phase invertor like in the Blues Express (EL84) version of the Trainwreck amps?

Look at version #8 on this:

http://www.el34world.com/charts/Schematics/files/Misc/trainwreck_express.pdf

OR ........... are you planning an LTPI version.

I'm open to trying to help but not sure what phase invertor you're planning to use.

Not sure what you're wanting to do?

With respect, Tubenit

[jjasilli]

1.  I agree with tubenit.  The trainwreck has cascading gain stages, after the tonestack.  The tone recovery stage can overdrive the the next stage; that stage can then overdrive the LTP PI.  This has signifigant consequences for overdrive tone. Due to the load line, a tube goes into cut-off at different points for the (+) vs.(-) signal curve.  AND signal changes phase for ea gain stage.  So building-up an overdrive tone over 2 gain stages allows the designer to sculpt the curve shapes which form the overdrive tone.  In this simple topography, the cathodyne phase inverter precludes this. 

2.  Probably the simple solution is to place a (switchable?) gain stage before the existing 1st stage in the PR. 

[Tone Junkie]

But after the new stage comes back before it hits the next stage you just bleed x amount of signal of to ground so it doesnt over power the next stage with any were from 330k to 220k to ground will solve your problem at least it works for a marshall type amp it also takes care of the big volume jump you get when adding stages. Bill

[topbrent]

You might consider the Mesa Lonestar style of adding an additional gainstage.  

It puts the extra gainstage in a different location (in front of the tonestack) in the circuit as compared to an Trainwreck Express (after the tonestack), but this schematic might give you some ideas.

It is nice being able to switch the extra gainstage in or out and have some control over the level.

This method should keep the reverb happy, also.

[VMS]

Something like this, maybe?

[tubenit]

I think VMS is on the right track.  I would suggest having that be switchable much as Geezer and I are doing on the TOS or Hoffman does on the hotswitch mod.

I think you'll need to tweak the component values probably. I would probably try a .0047 cap and a 100k or less dropping resistor as a starting place.

With respect, Tubenit

[jjasilli]

At this point, why not convert the PI to LTP!!!

[phsyconoodler]

Thanks guys!!
  I want to use the cathodyne PI so I don't have to add too many tubes.I may convert the tremolo to a Vibro-champ instead and just use an LTP anyway.The existing master in this amp sucks as is and a Type 1 master may be better anyway.
  I kind of like the idea of adding it before the first stage and switching it in and out like a hotswitch.
 I don't know!My brain is not working well the past week with a nasty head cold and sinuses that are throbbing.

Tubenit,you may not be as knowledgeable as Sluckey or PRR but your skill with layouts is outstanding.
  VMS,that is pretty damn good!
The Trainwreck Version 8 looks cool.

[Tone Junkie]

I second the idea of switching it in before the tone stack .After messes with the sound a lot

[jjasilli]

FWIW Dan Torres ("Inside Tube Amps") has a circuit for a switchable pre-gain stage - before the usual first stage.  The sw is a switchable input jack.  So, 1 input jack by-passes the pre-gain stage.  Plugging into the other input jack enables the pre-gain stage.

I think we're all wondering, with a number of Princeton builds under your belt, maybe you're thinking of branching-out to a different topology.  If so, then as pointed-out in above posts: cascading gain stages; where to put the tone stack; and PI type, become primary design considerations.

[phsyconoodler]

Yeah I'm thinking of adding it before the existing first stage and using a separate jack like a 2204.I have a raw control on this amp so I could convert it into a gain control rather easily.Kind of like using a pedal in front of the amp.Feed in a gain stage and control it with the raw control.I call mine a 'growl'.Then indeed it would be a growl!
  You guys are the best! Now I just need a layout so I can visualize it in my pea-brain.

[tubenit]

Now I just need a layout so I can visualize it in my pea-brain]Now I just need a layout so I can visualize it

I will try to help with that. However, it may not be til the wkend or later that I can get to it & it will be drawn with paralleled turrets like a HiWatt type board. Still that can be useful to reference even if you use another approach.

Here is a PR layout that could be modified IF someone wants to work on this prior to my getting to it.


With respect, Tubenit

[bruno]

this is actually quite nice, amps with extra gain stages before the tonestack seem to behave quite differently from amps with extra gain stage after the tonestack. No matter how much gain you dial into the ones with gain stages after the tonestack, they always have a clear/clean quality to the notes, in my experience.

[tyru007]

If your willing to dump the vibrato circuit, the extra triode could then be used as a gain stage adding no additional tubes to the circuit.  This would also free up a lot of room on the typical PR layout.   Probably add it after the reverb recovery stage and before the PI.   In this position it would have no impact to the typical fender reverb circuit.

[DummyLoad]

add a tube. 


my take on the question... two proposals depicted; both overly anal and overly complex. as usual.   

 


 

[Fresh_Start]

ISOTone - I'm not sure that you want to switch the NFB injection point to the added gain stage.  If the added gain stage is intended to introduce additional distortion (of a good kind), then putting it inside the NFB loop defeats the purpose (unless I'm missing something major).

Physco - How about "stealing" one triode from the reverb for your extra gain stage?  I'd leave the PR reverb circuit pretty much alone and use a 12DW7 for both driver & recover.  Use the low gain/higher current triode for the driver ("Section 2" pins 1, 2 & 3) and the higher gain/lower current triode for recovery ("Section 1" pins 6, 7 & 8).
http://www.mif.pg.gda.pl/homepages/frank/sheets/106/1/12DW7.pdf
You probably have to experiment with the cathode resistor for the driver and neither data sheet I found shows "typical" values for it.

As far as location in the circuit of the additional gain stage, the Trainwreck circuits seem to have the "extra" gain stage after the 1st stage/tone stack/second stage.  Since that's the topology you are going for, that's what I'd do.  The question for me is whether the extra gain stage goes before or after the reverb.  My initial thought is to try it before the reverb so that you are reverberating the signal going to the PI.  That would also help in terms of having reduced the reverb driver stage by stealing a triode there.  However, the other way might work great too.

Without time to draw this up, here's a possible topology:

first gain stage
tone stack w/ volume control
second gain stage
overdrive volume
third gain stage
split to reverb and 3.3 meg mixing resistor
PI
power tubes

I'm sure you'll have to experiment with the reverb mixing resistor too.  In addition, a "Dwell" control may be needed to balance the "Overdrive" control setting.  IOW you may need to reduce the signal going to the reverb circuit when Overdrive is turned up past a certain point.

Hope this helps,

Chip
power tubes

[phsyconoodler]

Actually I'm thinking of introducing it before the first gain stage,almost like a pedal in front of the amp.i think it would be the least intrusive and would not affect the reverb or anything else in the existing circuit.I know I would have to play with values to get the right amount of gain but I'm thinking it may be the way to go for me without redesigning the circuit too much.
  Just a thought..........

[DummyLoad]

If the added gain stage is intended to introduce additional distortion (of a good kind), then putting it inside the NFB loop defeats the purpose (unless I'm missing something major).

a signal clipped by any stage within a feedback loop looks less like the source signal, when that clipped signal (now NFB) is summed at the NFB entry node the resulting summed signal resembles the source signal even less, and thus will exacerbate distortion - the end result is that with NFB you accelerate the amplifiers transition into an over-driven state. NFB needs to be 180deg out, that's why i switch NFB to the stage closest to the concertina.

my intentions with the proposals presented, in the 1st schema, to overdrive the PI, output tubes, and  transformer. the second schema overdrives last gain gain stage et-al, and assumes the role of the summing amp. my intentions were also to leave the original preamp and reverb loop signal processing relatively unchanged. lastly, thoughts were that addition of a gain stage to overdrive the original first stage would negatively affect tone and cause all sorts of general nastiness within the reverb loop. both options allow bypassing the added gain stage. if i were to build, i'd choose the second.

the second proposal has an error - grid of last gain stage should be tied between 250K trim pot and 220K.

--ISO

[tyru007]

The "trainwreck" gain stage is a cold biased (is that right?) gain stage with the 10K at the cathode which results in assymetric clipping distortion at low signal level.  There is not much increase in gain as the cathode has no bypass cap.  Adding prior to the first stage will add some gain but it will not give you that distortion trademark sound since there is very little signal at that point in the amp.  

If your goal is to increase the gain a bit in the PR, then use a pedal or add a solid state boost or change some of the values in the circuit (i.e. reduce the size of the 3.3M resistor, etc.)

The PI in the princeton already tends to overdrive at high volume/gain with typically unpleasant results.   If your intent is a PR with a lot mote preamp distortion as you roll up the volume, then you need a stage that distorts the signal quickly without increasing the gain of the signal located at a point in teh amp to drive teh triode into clipping.  This would likley be after the second stage prior to the reverb or after the reverb recovery stage or as Isotone has show, as the recovery stage.  

Like mentioned before, you would need to figure out what to do with the NFB if after the recovery stage.

The Vox nightrain is a similar circuit but without NFB, reverb and vibrato.

[tyru007]

Isotone:  Do you think there would be enough signal to drive the gain stage into distortion in option 2?  

If were talking adding it in and not switching it in, I would vote for option 1.

[phsyconoodler]

So two triodes maybe paralleled into the first stage wouldn't be enough gain? Or maybe cascade the extra tube?I'd use my existing Raw control for a gain.
  Or maybe just build a tube preamp into a box and insert it into the circuit at the input?I just don't like the idea of disrupting the existing circuit for preamp gain.Just my hangup I guess.

[DummyLoad]

no disrespect intended but couldn't you just use a pedal?   

just about any reconfiguration will probably work with varying degrees of success, however, what sounds "good" varies between individuals.

the || triode arrangement with mix would probably sound really good, but i think one would lose that fender "chime" of the princeton pre-amp - the real question is this: what tone do you seek? do you want to preserve the princeton's tone as much as possible or are all bets off when the mods are made?

e-x-p-e-r-i-m-e-n-t !!!

topics such as this are the very reason why buttery and i built bread-boards for thermionic ckts. ideas can look really good on paper, but in the meat world they can [and usually do] end up sounding like crap.

a friend recently asked me to look into building him a princeton reverb, except he wants a 4 x 6V6 output configuration. he also expressed interest in additional gain and probably why i'm being a bit more verbose than usual. i'll probably strap up a PR on the bread-board again in the next few days and give some of the mods in this discussion a spin around the block. 

 

--ISO

[phsyconoodler]

Quote:"no disrespect intended but couldn't you just use a pedal?  "

  Of course I could 'just use a pedal'  but I find that properly implement tube gain sounds better than a solid state pedal.I could strap a pedal inside the cabinet and hook it into the input too,but I'm old school.I can't bring myself to do it

[topbrent]

What direction did you decide to go with this one?

[phsyconoodler]

I haven't decided yet.Still waiting for more input on this.

  I'm thinking of one extra tube before the first gain stage and blending it in with the Raw control uses a a gain control instead.
 I'm still not quite sure how to implement it,but I'm sure it would work.
   I bet with the right gain control it would be able to add some serious fatness to the existing preamp and then some insane gain if needed.The key would be how much is too much.
  The trainwreck adds it after the tonestack but before would be a lot like a pedal in front of the amp.Too much gain would be bad but just the right amount would likely be cool.At least that's what I think.

[phsyconoodler]

Here's my thought: use a 2204 arrangement,with one gain stage connected to pin two of the normal PR amp and the input connected directly to the gain pot with the 470k/500pf treble peaking marshall idea and simply coupled through the plate resistor by a cap.
  I think that would work just fine,but how much gain it would add would be regulated by the cathode cap/resistor.
What are the thoughts on this?

[topbrent]

This is a Allen 5F1+, which is a high power 5f2a with a cascaded gain stage in front.

You could probably take this multi input concept and adapt it to a traditional ab763 style front end.

[phsyconoodler]

That is exactly what I was thinking about,except I don't want to use the high/low input jacks and just use the high.Then I would simply use the volume control as the 'gain' control and the amp already has a master.This is basically the same as the 2204 marshall with no treble peaking circuit.
  I may or may not use the treble peaking,we'll see what it sounds like first.

Thanks a bunch Topbrent!!