New build: Princeton Reverb with mods - advice wanted!

[klooon]

On a bit of perf board

OK, and then mount that to the eyelet board. Good idea.

[klooon]

OK, I am almost there with the schematic, so I can order parts now. I will order a bunch of extra resistors (especially 100k and up), switches, pots to tweak. MM Transformers are already in. Deluxe Reverb chassis is on its way... I am planning to use carbon film (1W) resistors (they are in my 18W build and they are nice and sooo quiet). Further, I will get polyester Mallory caps and CTS pots.

Two more questions:
- is there a way possible to avoid a switching input jack as the switchcraft jacks seems to continu failing on me after a while.
- which diodes should I use exactly in the tremolo circuit and the biasing circuit?

[tubeswell]

- is there a way possible to avoid a switching input jack as the switchcraft jacks seems to continu failing on me after a while.

Not if you want to keep the noise quiet when the cables are unplugged

- which diodes should I use exactly in the tremolo circuit and the biasing circuit?

1N4007 in the bias circuit

12V zener between the mosfet source and gate

[klooon]

OK, here is the first version of a possible layout. I have integrated the Excelsior tremolo circuit. Further, I have decided to use the stereo jacks for the speaker outputs.

I am not fully sure about how I wired some of the pots (voltage divider vs. variable resistor)...

Any comments are welcome!

[klooon]

Replace the 6.8K with a short. If you use a 100K Audio taper pot, you will have useful mid adjustment over the first half rotation and the second half will remove the scoop. The Bass and Treble controls will be less effective however. See if you can find a schematic for the Kendrick Texas Tea control. The New Joy Zee amp has it. It's another pot that creates a shunt across the treble cap. Another thing to try is a pot in place of the "slope" resistor. Duncan's Tone Stack Calculator will help you visualize what changes can do.

OK! I have been playing around with the Duncan Tone Stack Calculator to see what the tone controls do, including the tweed/texas tea control. It seems (correct me when I am wrong, please) that the 100k Mid pot takes you from a deep mid notch to flat EQ. The frequency of of the mid notch does not change so much and the treble and bass controls become less effective:

Figure 1: Figure 2:

Then I changed the value of the treble cap and you can see that the mid notch dissappears, but also the frequency of the notch shifts to lower values.
Figure 3: Figure 4:
Note that the bass and treble controls are still quite effective.

Lastly, I tried to simulate a tone-stack bypass by putting the mid resistor to a extremely high value (100MOhm).
Figure 5:
You can see that you get a total boost and further the (tone) controls become uneffective.

To me the Texas tea or tweed control is a combination of a mid boost in Figures 3 and 4 and the tone-stack bypass (boost) of figure 5. Could this be right?

So it's either using the mid control which results in a mid boost without the shift in 'notch frequency' or a mid + total boost with a shift in notch frequency to lower values.

To me if I want to get to a more tweed like flatter EQ then the mid control option would be the best (more bass is preserved on the way to the flatter EQ).

Changing the slope resistor to a lower value results in a shift of the notch frequency to the right, which is maybe a good thing as well to get a more Tweed EQ!!?? Here I used the 56k value to compare:
Figure 6:

What do you think? Would the mid control be the way to go? Anyway, both the tweed control AND the mid control is too much?? Even the tonestack by-pass option is maybe too much as an option...

[loudthud]

The Texas T control might pass slightly more signal when it's all the way up. It's disadvantage is you have to add the pot and knob. I would just build the amp with the 100K mid pot and experiment with the Texas T too see if it adds anything useful.

[klooon]

The Texas T control might pass slightly more signal when it's all the way up. It's disadvantage is you have to add the pot and knob. I would just build the amp with the 100K mid pot and experiment with the Texas T too see if it adds anything useful.

Or I could put the Texas T control instead of the mid pot?! Do you think the tonal characteristics are much different? Am I right assuming that the tweed has a mid notch at higher frequencies as the BF? Going for the Texas T is then going in 'the wrong direction'??

[tubeswell]

A James stack could also work very well in a PR. The attraction of a James stack is a flat response when both pots are centered. The treble control also has a really impressively wide bandwidth response. (To me tone controls are all about getting as much change as you can for your money. YMMV)

[klooon]

A James stack could also work very well in a PR. The attraction of a James stack is a flat response when both pots are centered. The treble control also has a really impressively wide bandwidth response. (To me tone controls are all about getting as much change as you can for your money. YMMV)

Really impressive! The controls are really sensitive:
Figure 7:


The thing that holds me back is the level (insertion loss) at flat response (more than 20dB 'loss') and the notch is shifting quite a bit to higher frequencies. I already ordered the parts for the current PR stack (with mid and maybe then the Texas Tea), so I think I will keep that direction for the moment.

[klooon]

I have revised the layout a bit, especially the tremolo circuit (also found some "missing pieces"). I hope it's fine now?! Also some advice on the grounding scheme is more than welcome!

[klooon]

Here is a new schematic of the amp:


I omitted the Texas Tea control (for now) and with the stereo jacks for the speaker multiple outputs.

Here is a close up of the Fender Excelsior tremolo circuit:


Further, I adapted the layout. Now version 03:

- corrected one error in the tremolo circuit (Mosfet connection to 'lower side' of the 220k)
- I put the OT on the drawing at a different, more tidy spot.
- different grounding scheme, based on the ideas here:
New Tremolux 5G9 build --> input welcome!

An extension of the philosophy of separating high current and low current grounds is one where you try to keep currents in each block of circuitry within that block. When currents need to flow between blocks, (Signals and DC) then only those currents should flow between the blocks. This requires you to know where the current flows. The so called star ground supposes that since the grounds all come together at one point, only the ground currents for each circuit or block would flow to the correct circuit. The problem is that a stack of spade lugs on a transformer bolt makes far from a perfect star. It would be better to have a small eyelet board with one eyelet in the center and a circle of eyelets around it. A short wire from each eyelet in the circle would go to the central eyelet. Then a single wire from the central eyelet to the chassis.

At least, I hope it's a bit better now?...

[klooon]

I have adapted a few more things:
- connection pin2 to pin7 of V2 was missing
- in the reverb circuit, the link from the 3.3M and 10pF to the point inbetween the 500pF and 0.02uF was missing and there was a connection from the 500pF to 470k which should not be there.

Here is layout version v04:

[klooon]

Another update: version 5 of the layout:


As KOC teaches: troubles with star grounding:

Star grounding in a guitar amp is generally a bad idea. This brings currents together that should never see each other and you end up with more hum than a randomly-grounded amp.

I have adapted the layout (thanks LT):
- I have made 4 local star nodes: 1. Power section 2. Tremolo section 3. Reverb section 4. Preamp section.
- the reverb cathode is connected to ground in the preamp section.
- reverb and dwell knob grounds go separately to star node.
- fill heating CT (at V5) doesn’t go to ground in the cathode bias switched situation, so I put the 100R’s at the pilot assy and ground them to the first star node.
- regarding the Mosfet: both the pinout was not okay (TO220: D is in the middle) and the Zener. I was confused with the internal Zener of the IRF820: the external protecting 12V zener goes from Source to Gate.

[überfuzz]

As KOC teaches: troubles with star grounding: ETC...

The layout looks very Fenderish. Ever considered a Fender grounding scheme?

[klooon]

The layout doesn't only look very Fenderish, it is very Fenderish. For a Fender grounding scheme I probably need a much bigger soldering iron for soldering to the brass back plate and chassis. And I should put some (other people's) thoughts in the grounding scheme anyway, because of the mods that I have in mind.