princetn.txt PI Power Supply mod for Princeton Reverb amps * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * From Steve A. (AMPAGE Tech Talk BBS 1/26/00) Get a schematic for the Princeton Reverb from the FFG site or the Pittman book and look at the power supply. The schematic I am looking at shows two nodes of the power supply labeled as "B" and "D". Between these two nodes is an unused point which would presumably be "C". You need to move the wire from the PI (trace it from V4P6) normally going to point "D" over to this unlabelled point, "C". Leave the other wires connected to point "D". * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * From John Stokes (AMPAGE Tech Talk BBS 1/27/00) Practically speaking, the wires for B+ distribution on the eyelet board are daisy-chained. Sadly, the wires from the other preamp tubes go to the eyelet that feeds the PI, which then has a wire to the filter cap. This means you'll have to remove the wire that feeds the other preamp tubes, and extend it over to the filter cap separately, into the old location at the end of the voltage divider. Just move the wire already there to the next higher tap to take care of the PI. Hope this makes sense. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * From Mike Tolomeo based on suggestions from John Stokes (June 1998): Increasing the plate voltage of a split-load phase inverter A little more volume can be gotten out of Princetons and Princeton Reverbs by raising the plate voltage on the split load phase inverter. In Princeton Reverbs there is an untapped filter stage that is ideal for this purpose. By moving the plate supply for the phase inverter back to the unused filter stage, the plate voltage changed from 195v to 250v. The higher voltage gives the phase inverter more headroom. This is not a tremendous difference, but it may allow the phase inverter to drive the power tubes closer to saturation. And, unlike converting to a long-tailed pair phase inverter, it doesn't change the variety of tones that you can get out of this amp. Locate the wires marked in blue on the diagram (they won't necessarily be blue in the amp). Unsolder them from the end of the 56K resistor, lift them off the board, solder them together and heat-shrink them. [Unable to display image] Now connect the violet wire to the end of the 1 watt 18K resistor at the top left of the fiber board to the 56K resistor where you just removed the 2 blue wires. [Unable to display image] Recently there was a post on alt.guitar.amps that warned about possible excessive voltage on the cathode when doing this mod. The poster said that a 12AX7 tube could take up to 100v on the cathode but not more. His schematic of a Princeton listed 65v being present on the phase inverter's cathode in the stock design, and thought there would be enough headroom so that the 100v max wouldn't be exceeded. I didn't measure the PI's voltage before doing this mod, but my schematic listed its voltage at 50v. It's currently at 65v after doing this mod, so it doesn't appear that the tube is being stressed. *** This article can be viewed and downloaded with the drawings as: http://www.ampage.org/blueguitar/prtweak.pdf * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * More from John Stokes: The PR is rated at 12 Watts. I was wondering why that was, since it is so very similar in output tube compliment and voltages to the 20 Watt rated Deluxe Reverb. Moreover, a friend has a Super Champ (another 2x6V6, single 10") amp that is damn loud. Hmmm...... I found out what the deal was. When driven hard, I noticed severe asymmetry in the output waveform. One side of the output was clipping much sooner than the other. Why?? Further probing revealed that one of the 6V6s was not even being driven fully into saturation!! I then started looking at the PI, which in that amp is a split-load type. It turns out that the PI was running out of dynamic headroom before it could drive both sides of the output to full tilt. I then decided to try to remedy that by increasing the B+ voltage (and hence the headroom) available to the PI. This was easily accomplished by moving the B+ feed to the PI up one notch on the power supply voltage divider. Problem solved. The increased voltage allowed the PI to fully drive both 6V6s. The output was not much more symmetrical under drive, and the measured output power jumped to 18 Watts. All this by essentially just moving a couple wires from one place to another. Another thing I noticed is that the coupling cap into the PI was a .02. The input impedance of a split-load PI is VERY high, since the 1 meg grid return resistor is returned to the top of one of bottom side load resistor, and due to the degeneration, the effective input impedance is increased considerably. Anyway, this provides for way too much low end response, leading to the often mentioned "Fender fartyness". Drop that cap down to .001 and the farts get a dose of gas-x. The stock speaker is a rinky-dink piece of poop. (Mine came with an Oxford 10J4.) The C/P10 Q should work well in an amp that's had it's output boosted. These tweaks make the amp actually gig-able. Mine is not harsh in the least. I'm using the JAN Philips 6V6s, and they're fairly smooth under overdrive conditions. As always, YMMV. John Stokes"