The Dirty Princeton Reverb

[Ledzepp007]

I had an idea to do a Princeton Reverb with modifications inspired by the Tweed OD Lite (awesome) and the Friedman Dirty Shirley. I really have very little clue on how to implement reverb (I've never tried). Attached is the first draft of my idea. Would y'all think this would work and sound all right?

Thanks!

[tubenit]

Some suggestions to consider.

With respect, Tubenit

[tubenit]

Honestly, since there is no relationship to a Princeton Reverb in the topology .......... AND since it will be a fairly high gain amp,  I think I'd be
inclined to build it with an active FX and use delay instead of reverb.  I tend to prefer delay over reverb for higher gain and reverb over delay for
lower gain.

OR .......... you can use reverb and delay pedals individually or together in the active FX.

You could actually lose the master volume and simply use the FX pots as a master volume.

With respect, Tubenit

[SnickSound]

So effectively this is a single channel JCM800 preamp into a cathode biased 6V6 power amp, looks a lot like this amp I built:


If you ignore the more complex preamp (it allows going between JTM45, JMP and JCM800 preamp specs), you can see how I inserted a defeatable tube FX loop between the treble wiper and master volume.  As tubenit mentionned, this will give you more flexibility. Since this amp will have a decent amount of gain in the preamp, you want to add time based effects as late as possible

This is a cool amp with a surprising amount of overdrive due to the lower voltage compared to a full blown JCM800.

More info about my projects (plus link to my Youtube channel if you want to hear it in action): http://www.snicksound.com/p/15w-6v6-plexi.html

PS: You'll notice diodes around the cathode follower stages, these prevent the grid from being more positive than the cathode, a very cheap protection that will prevent tube meltdown. With this in and a GZ34 rectifier, I can't see a reason to have a standby switch.

[Ledzepp007]

Wow. This is some awesome info. Thanks guys. I do like the Effects Loop idea. Although I’m not opposed to the MOSFET CF driving the tone stack, I would have four preamp tubes to play with.

I’m wondering if my filtering is a little too stiff. I do want higher gain, but with a little bit of play in the supply (obviously not enough that the amp becomes unstable, though). Would 40-20-20-20-16 work? That way, I could use the JJ cap can with a 16uf on the main circuit board.

[SILVERGUN]

Wow. This is some awesome info. Thanks guys. I do like the Effects Loop idea. Although I’m not opposed to the MOSFET CF driving the tone stack, I would have four preamp tubes to play with.

I’m wondering if my filtering is a little too stiff. I do want higher gain, but with a little bit of play in the supply (obviously not enough that the amp becomes unstable, though). Would 40-20-20-20-16 work? That way, I could use the JJ cap can with a 16uf on the main circuit board.

40-20-20-20-16 should be fine. The only issue with using the cap can is that you can't separate the ground connections, which doesn't matter so much on a stock PR, but could be problematic on a high gainer.

If you wind up opting for just reverb I would suggest separating the R and R2 tap points. By moving the R point earlier in the preamp you'll get a cleaner signal for the reverb that skips around the late gain stage and isn't affected by the Master vol. setting. You could still mix the reverb return in where R2 is but you'll have to experiment with values to get it right.

My schematic edit was just a quick rendering to visually explain the concept. Very quickly drawn and obviously not tested.

[SnickSound]

Wow. This is some awesome info. Thanks guys. I do like the Effects Loop idea. Although I’m not opposed to the MOSFET CF driving the tone stack, I would have four preamp tubes to play with.

I’m wondering if my filtering is a little too stiff. I do want higher gain, but with a little bit of play in the supply (obviously not enough that the amp becomes unstable, though). Would 40-20-20-20-16 work? That way, I could use the JJ cap can with a 16uf on the main circuit board.

After some testing, I decided to go with 64-32-32-32-32 (not in a single can) in my amp.
You can hear the tests here:

There's more to it than the capacitance and the amp just seems more responsive with 64uF. With 32uF it sort of didn't keep up with me, so to speak.

As others have said, a single can is not the best idea, especially with something that has decent preamp gain like this. Mine has 2 dual 32uF cans (one is paralleled for the main 64uF reservoir cap, the other feeds the screens and phase inverter)), and a further dual 32uF cap (not a can type, one that goes on the board) for the preamp. Oh right, I decide to run the FX Loop of it's own supply so there's an 16uF for that tube (node is paralleled with phase inverter node for the highest voltage and thus most headroom)

[Ledzepp007]

What about this for a layout for the effects loop?

[Ledzepp007]

Wow. This is some awesome info. Thanks guys. I do like the Effects Loop idea. Although I’m not opposed to the MOSFET CF driving the tone stack, I would have four preamp tubes to play with.

I’m wondering if my filtering is a little too stiff. I do want higher gain, but with a little bit of play in the supply (obviously not enough that the amp becomes unstable, though). Would 40-20-20-20-16 work? That way, I could use the JJ cap can with a 16uf on the main circuit board.

After some testing, I decided to go with 64-32-32-32-32 (not in a single can) in my amp.
You can hear the tests here:

There's more to it than the capacitance and the amp just seems more responsive with 64uF. With 32uF it sort of didn't keep up with me, so to speak.

As others have said, a single can is not the best idea, especially with something that has decent preamp gain like this. Mine has 2 dual 32uF cans (one is paralleled for the main 64uF reservoir cap, the other feeds the screens and phase inverter)), and a further dual 32uF cap (not a can type, one that goes on the board) for the preamp. Oh right, I decide to run the FX Loop of it's own supply so there's an 16uF for that tube (node is paralleled with phase inverter node for the highest voltage and thus most headroom)

Holy crap, what a cool comparison. I was thinking I might run two cans, one 50-50 for the reservoir and screens, and then a 40-20-20-20 for the PI, effects loop, and preamp. 40 for the PI and 20 for each the loop and each of the first two AX7s.

[Ledzepp007]

Here is a layout draft. Is the FX Loop implemented properly?

[SILVERGUN]

Which schematic did you wind up using?
It is always helpful if you post the schematic along side of your layout.
One other suggestion I would have would be to use the same colored wires throughout your layout to represent plate, grid and cathode connections, if possible.
It makes things easier to trace. Some of us carry out that same concept throughout our build and I tend to use red for plates, white for grids and black for grounds.
The actual colors don't matter but sometimes just the fact that things are colored in a uniform fashion makes finding a mistake easier.


Once we see the schematic we can then determine if your effects loop is implemented properly.


-Did you really intend to switch between a 10K or 12K cathode resistor on the second stage?
-Did you intend to connect the plate to the grid on the second preamp tube?


Im voice typing on my cell phone, so that's all Ive got for now but hopefully this will help you get some help.

[Ledzepp007]

Thanks for the reply. I used tubenit’s schematic from above (Dirty Princeton Fx) with a couple of small mods. One of this is using a 12k bias resistor with a .68 bypass cap on the higher gain setting from the switch instead of 2.7k. The 12k with the cap still gives more gain than the 10k with no bypass cap. I did mean to connect the plate to grid on V2 as it is the cathode follower/tone stack driver. I believe that I followed the schematic correctly in my loop layout, but not 100%. Am also not sure about the switch to activate/deactivate the loop. I just used a simple SPDT switch. Would there need to be a mixing resistor involved between the loop and the LTPI?

Thanks!!

[SILVERGUN]

Thanks for the reply. I used tubenit’s schematic from above (Dirty Princeton Fx) with a couple of small mods. One of this is using a 12k bias resistor with a .68 bypass cap on the higher gain setting from the switch instead of 2.7k. The 12k with the cap still gives more gain than the 10k with no bypass cap. I did mean to connect the plate to grid on V2 as it is the cathode follower/tone stack driver. I believe that I followed the schematic correctly in my loop layout, but not 100%. Am also not sure about the switch to activate/deactivate the loop. I just used a simple SPDT switch. Would there need to be a mixing resistor involved between the loop and the LTPI?

Thanks!!

Ok, I scanned over your FX layout and it looks good. It has been brought to our attention recently that we might want to be using a 25K pot for the send so as to retain the low output impedance of the cathode follower, so there's that. 
You're missing the 120P cap from grid to ground on the recovery....intentional mod?
 
The reason I asked about the connection of plate to grid on the second preamp tube (V2) was to get you to look at it harder. It's perfectly fine to make that connection, but the plate supply marked "X" shouldn't be going to your ground bus. See attached for where to move it.

I didn't have time to check the whole thing over, but I would still suggest modifying tubenit's schematic so that it matches your plan. This way, the guys won't have to guess what your thinking.

I like the plan.

I don't have time to review the FX switching but it doesn't look right to me. I have used the DPDT relay that Doug sells to perform this function and there is plenty of info on the boards about implementing a switchable loop. I'd go back over your idea.
There doesn't need to be a mixing resistor added to T's schematic, it works as shown. It is drawn so that (if you use a switch) it gets placed in series with the signal when it is switched in, and removed from the signal path when off.
So, you need to switch FX{open-closed}FX and FX2{open-closed}FX2....there could be another way, but that's the way I've done it.

[SILVERGUN]

For the V2 cathode follower hookup there is a simpler (very common) method of bringing the B+ supply over to the tube at pin 6 (X).
Then connect the 100K plate resistor right across the socket to pin 1 and tie that over to pin 7.

[SILVERGUN]

So, I was going back over this because I was going to suggest that you use at least one grid stopper and I noticed that you do show a 470K grid stopper on your layout (see attached).
The problem with that is that it doesn't match T's schematic and that is a big problem because it was designed to be a voltage divider in conjunction with the 470K that is going to ground. By moving the resistors around you have eliminated the voltage divider function.

All Resistors Matter.

[SILVERGUN]

You need a ground connection here:

[Ledzepp007]

So, I was going back over this because I was going to suggest that you use at least one grid stopper and I noticed that you do show a 470K grid stopper on your layout (see attached).
The problem with that is that it doesn't match T's schematic and that is a big problem because it was designed to be a voltage divider in conjunction with the 470K that is going to ground. By moving the resistors around you have eliminated the voltage divider function.

All Resistors Matter.

Good call, yes it does eliminate the voltage dividing function. I will fix that. Nice catch. I should have noted the ground on that pair of 220k resistors.

[SILVERGUN]

So, I was going back over this because I was going to suggest that you use at least one grid stopper and I noticed that you do show a 470K grid stopper on your layout (see attached).
The problem with that is that it doesn't match T's schematic and that is a big problem because it was designed to be a voltage divider in conjunction with the 470K that is going to ground. By moving the resistors around you have eliminated the voltage divider function.

All Resistors Matter.

Good call, yes it does eliminate the voltage dividing function. I will fix that. Nice catch. I should have noted the ground on that pair of 220k resistors.

I'm here to help. If your name was Loverboy007 I probably wouldn't care as much.

I looked at it again and I'm sure now that the reason your FX switch is giving me a headache is because you have mindmelded tubenit's tonestack / Master with SnickSound's switching and something got lost in translation.
If you could scratch that into a little schematic clip I will try harder to figure it out. I like what SnickSound did there, but it seems like you've got a wire crossed.

Either way, keep up the good work and I look forward to hearing a demo clip off Physical Graffiti when your done.

[Ledzepp007]

OK, I had five minutes to update the layout based on all of y'all's excellent feedback. Check it out if you want!

[Ledzepp007]

Here's a fixed-bias version (attempt). Notice any errors?

I also switched over to Cliff Jacks (I wanted to insulate the jacks).

Thanks.