Princeton Reverb

Here's the first tube amplifier rebuild article I wrote. I thought I'd include it as part of this website because it too is instructive. It is amusing to me to read my own words from 1999 because I have really advanced my tube amp knowledge since this was written. As of this writing, in the fall of 2010, I still play and very much enjoy this amp. The original text starts after the picture.

This is a 1971 Silverface Princeton Reverb, which I purchased at the Great American Guitar Show in King of Prussia, PA (near Philly) in July 1999. For those of you who don't know, this is a 12-watt, all tube amp with reverb and tremelo. The front panel knobs are Volume, Treble, Bass, Reverb, Speed, Intensity. The speaker is 10" and 8 ohms. One channel, no effects loop, no speaker emulation. Remember, simple is good! The amp was very clean, and had all original parts, except for several of the tubes. The amp sounded pretty good at low volumes, but it clearly needed a cap job, new tubes, and a new speaker; and there begins an odyssey of restoring and improving this vintage piece of amp Americana.

The following section details all the steps I took to achieve tonal nirvana with this simple amp. Even though it is a bit long-winded, I wanted to take the time to put in all the detail of what it took to restore and improve this fine amp so that people that haven’t gone through this process can understand what is involved, and what it will cost.

Restoring the amp Part I: My 7 hour Tube Amp Lesson

I like to take tube amp lessons. Basically what this means is that when I have an amp that needs work, I take it to my amp tech, Rick, who graciously lets me watch him work, and patiently answers all of my questions. Yes, I slow him down, but it is a great way to learn.

I did a fair amount of research about what I wanted before taking the amp to Rick’s house with the assistance from the good folks on the alt.guitar.amp newsgroup. Here’s what I decided I wanted:

1. New tubes and bias
2. Cap job
3. Resistors as needed
4. Replace two-prong with three-prong cord
5. Ground the courtesy outlet
6. Disconnect the three-way ground switch and use it to turn the tremelo on and off.
7. Replacement speaker - Webervst C10Q
8. B+ Phase Inverter Mod
9. Install screen grid resistors to protect the power tubes
10. Install a bias adjustment pot
11. Balanced phase inverter (Rick’s suggestion)

Bear with me if I don’t have all of the terminology correct, or a completely clear description. Remember, I’m taking lessons, not giving them.

Anyway, I got to Rick’s house at 2:00 pm on Sunday afternoon (by appointment) and I woke him up. He was out late doing sound for a local band. He lives out in the woods and has a nice, quiet property with a cat and a house broken rabbit.

I lugged the amp down to his basement workshop, which is strewn with ex military parts, lots of oscilloscopes (for parts), amps, tubes, and various and sundry stuff. His workbench is in a corner of the basement, surrounded by shelves. Facing the workbench are all types of electronic equipment and reference manuals. Rick used to buy and sell military surplus, so he has a lot of nice test equipment.

I don’t know what all of it is, but I know there is a 20 amp variac, two meters to show house power and variac power, a signal generator, various expensive looking meters, soldering irons, test cables, compressed air, a patch bay, and a tube tester.

The floor is covered with a thick rubber mat with holes in it. It is comfortable to stand on which is something I came to appreciate over the next seven hours.

The first thing Rick did was remove the chassis from the cabinet, and put it on two wood blocks designed to hold it in place. He then drained the capacitors by hooking up a wire and alligator clip to what I think was the main capacitor (metal can) and another to the chassis.

He then looked it over for a good while, testing solder connections, and generally looking around. He pointed out to me the chassis was stamped 9/10/1971 and had the assembler’s name stamped as well. His name was “Castillo”. By the way, the serial number of this amp is A25810.

He also pointed out that the power transformer was unusual. Each half of the primary windings has two wires (not one) coming out of it. It is stamped with the words CSA Test, the number 010021, and the number 36 or 30 at the bottom. Rick speculates this was a test transformer that is stronger than a PR but less than a DR. Can anyone shed light on this?

Next, he wanted to hear how it sounded so he plugged it into the variac, plugged the speaker back in, and brought it up to 120 volts. His test guitar is a Telecaster which from the ceiling behind him. He turned the volume, treble and bass up to 10 and tried a bunch of different strums, patterns, and harmonics. I came to know these licks very well by the end of the day since he played them quite often. By the way, Rick isn’t a guitar player per se; he basically uses it to test amps.

After playing a bit, he had a pinched look in his eyes and I knew he wasn’t that pleased with what he heard. Frankly, I wasn’t either. At lower volumes it was fairly sweet sounding, but at higher volumes it was harsh and farty.

He checked the speaker by hooking something up to it, and I don’t recall what it was, but he said it didn’t seem to have a lot of life left in it. The speaker was a Fender Special Design with no serial numbers anywhere on it. He replaced the speaker with the one I had brought along, which was the Weber C10Q 8 ohm that I had recently purchased. Before Rick put it in he looked it over and murmured how impressive looking a speaker it was. He loved the quality of the workmanship.

It had been an hour since I got there, and having just woken up, Rick was thinking about getting a pizza. I offered to go get one, but he settled for a banana from the kitchen. Back to work.

Since I had brought a full compliment of tubes, he wanted to test them. I had some Vissueaux 6V6’s, a Phillips 5U4GB rectifier, a Sovtek 12AX7LPS, some GE JAN 12AX7’s, and a Phillips 12AT7. After selecting the two preferred 6V6’s, he installed all of the tubes in the amp and brought it up again on the variac. Out came the Telecaster, and boy did the Weber speaker sound good.

After powering down and draining the caps, Rick said he wanted to balance the power supply to the power tubes. I didn’t fully understand this, but he said that the existing circuit isn’t equally balanced. From the parts list I got at the end of the session, I believe he installed a 100yf/100v bias cap.

Next up was to bias the power tubes. This is where I asked him about installing a bias pot, and showed him Lord Valve’s recommendation from the newsgroup to install a resistor in line with the bias pot so that it couldn’t accidentally go to zero. Rick thought this was a good idea. He turned around to the shelves and began looking for the right bias pot. He had a bunch and he tested them one by one to find the one with the right amount of resistance. He wasn’t happy with any of them, and he opened another drawer and took out a Bournes 10 turn bias pot that he had taken out of an oscilloscope. It tested fine, and had quite a wide range.

This gizmo is cylindrical with a turn screw at the end of it. It is about 1” long and 3/8” in diameter. He installed it onto an existing bolt in the chassis using a plastic wrap around holder that fit over the bolt, with a nut attached, leaving room to adjust the bias with a small screwdriver. It is a very neat installation that didn’t require any drilling or silicone. I realized this installation gave the same kind of control as the front panel bias on the Speedster line of amps, which feature a 10 turn bias pot and a big meter on the front panel of the amp. Pretty cool.

With the aid of the 10 turn bias pot, Rick adjusted the bias, and it was possible to easily to control it to almost the .01 level. Down came the Telecaster off the wall, and the amp was really starting to sound better.

One of the reasons the amp lesson lasted seven hours is that Rick would take the time to show me how the amp sounded after each incremental adjustment so that my ear would get educated. It was a valuable lesson. One of the things I noticed during the course of the lesson is the sound would swing from clear to cloudy and back again, depending on the change. In other words, fixing one thing might muddy up the signal, which in turn would need corrected.

By now, three hours had gone by, and we were both starting to think pizza. We called in the order to the bar (the only place open in the sticks on Sunday) for a large onion and pepperoni pie. I volunteered to make the pizza run so I went upstairs and almost tripped over the cat and rabbit, which were racing around the house playing with each other. It was pretty funny to see.

I got to the bar at the requested time, but the pizza wasn’t ready for another twenty minutes. I guess they wanted a few extra bucks for drinks while you wait for your pizza. Since I was in the middle of a lesson, I had a coke. Then back to the house with the pizza, which certainly hit the spot. As we ate, we talked about the next items to work on. Rick said he likes to tackle the pre-amp side of things before doing anything more with the power side.

We discussed the B+ phase inverter mod, which I had originally seen discussed on the alt.guitar.amp newsgroup by Tremolux. The intention is to keep the amp from farting out by moving the B+ feed up one notch on the power supply voltage divider to give the phase inverter more dynamic headroom. Troy (from a.g.a) suggested an alternative method by reducing the value of the coupling cap feeding the phase inverter. I showed Rick the instructions that Tremolux had written, and the alternate methods of doing these mods as suggested by Troy.

Rick said that before trying these, he would like to work on the pre-amp side to get that sounding as good as possible. He also suggested it might be a good idea to get the amp sounding as good as it possibly could with the existing phase inverter before trying the mods. I couldn’t fault his logic, so that’s how we proceeded.

Rick tested a number of different voltages in the amp to see what capacitors and resistors needed changing. He was pretty quick at this, and he didn’t need to refer to the diagram. This was old hat to him, having worked on countless Princeton Reverbs. He told me that he was going to replace about 10 capacitors with Sprague Orange Drops. He said the main capacitor in the metal canister would likely need replacing at some point, but it wasn’t showing too much leakage now. I believe he said about half a volt. As you’ll see, this did end up getting changed in Part II of my amp lesson.

He proceeded to replace caps. He said he would normally replace them all before turning the amp on and testing it, but because I was there, he stopped after each section of the pre-amp and played the Telecaster so I could hear the incremental changes. By the time he got done, the amp was much tighter in the bass, and the harmonics were really sustaining beautifully. By the way, he didn’t change any of the blue coupling caps (see Part II). He said that sometimes changing them can too radically alter the sound of the amp, and he wanted to wait until later to see if it was needed.

One thing I noticed him doing all the time was being very careful not to let little bits of solder fall into the board. He used a hand-held device that sucked the solder out of the holes. It was a spring-loaded suction device. After taking out a cap or a resistor, he would heat up the remaining solder connection, and then trigger this device, which would suck the remaining solder up into it. Rick said he preferred removing the old solder and using new solder to assure a good connection.

Next on the agenda were the pre-amp tubes. With the volume, bass, and treble on 10, he tapped on the first tube, which I had selected as a Sovtek 12AX7LPS. It was pretty microphonic when tapped; much louder than the GE’s. Rick suggested I try one of the JJ’s in this position, which we did, and it was much quieter. I liked the tonal qualities of it as well. For the moment we skipped the 12AT7, and tried different tubes in the third and fourth positions. The bottom line is that I liked the GE’s in these positions. The JJ in the third position had a more complex harmonic structure after the initial note sounded, but I liked the attack of the GE better, so that’s what we left in the third position.

The reverb circuit was next. This amp has a beautiful sounding reverb, and I was quite happy with it before I brought the amp in. Rick listened to it and said it was a typical sounding Fender reverb, but he wanted to try a couple of things. He changed two capacitors on the reverb circuit, which made it sound a bit crisper with better lows and highs. Next he suggested trying a 12AU7 in the reverb circuit to tame it a little bit. I liked the change and it is still in the amp. It makes the reverb seem more a part of the note.

One thing Rick noticed was that the reverb was still audible when the reverb was turned off. There was some kind of leakage. I wasn’t real concerned about it because it was less reverb than I always use. Rick put the 12AX7 back in so the voltages would test properly.

About this time, two fellows from a recording studio on the other side of the county showed up on a whim. They had heard one of Rick’s customers had purchased a mid 1960’s Marshall head and 8 X 10 cabinet. Apparently the customer had driven to Ohio and spent $3,500 on the amp and now wanted it restored. At first I thought the head was a JTM-45, but it was a 1987 instead. He fired it up, and man-alive do the 8 10” speakers sound amazing. We then hung out a while talking about amps they like to use in the studio. A Super Champ was the latest addition and they were pretty excited by it. Since it was getting late, I asked Rick to let the reverb problem go until next time so that we could finish some of the other things.

Next on the list was reworking the AC section. He took the ground switch and capacitor out of the power supply, and added a long, thin, 3 prong cord that fit through the grommet without a problem. He grounded the chassis, and made sure the courtesy outlet was showing a good ground as well.

After the amp was grounded, there was less hum in the amp. Now it was time to put in the screen grid resistors and the grid stopper resistors. We talked about this for a while because Rick said it would affect the sound. We listened carefully to the amp before putting this mod in. There were gorgeous swells in the sound. You could hit a note and count four swells, with lovely overtones.

Rick said it was important to put these resistors right on the socket or it wasn’t worth doing. After the resistors were in place we listened carefully again. The sound was still great, but we’d lost one of the swells. It seems the Visseaux 6V6’s are allowed to do more of their thing, if the resistors on the socket don’t hold them back. On the other hand, those resistors are what protect the tubes and output transformer from a short. Since I want to be able to use this amp at band practice, and will need to keep it on 7 most of the time, and turned occasionally to 10, I decided to leave the grid resistors in place. Deluxe Reverb’s are wired this way, but it certainly speaks to the fact that in amp circuitry, simpler is better. I heard it quite clearly.

It was about 9:15 and I had been listening to changes in my amp’s tone all day long. It was really neat how tuned in my ear was at the moment. This came in really handy as we tackled the last project for the day, using the ground switch to switch the tremelo on and off. I really like to travel light when I go to practice, with the guitar in one hand (with spare tubes and fuses in the guitar case) and the amp in the other. One trip from the car, and that’s it. I don’t like carrying a foot pedal behind the amp, and I may only use the tremelo twice a night, and hence the reason for wanting to use the now unused ground switch.

Earlier, while Rick was working on the AC section, he asked if I’d prefer a two-way switch for this purpose. I did, and he replaced the three-way switch. I thought he would use this switch to connect to the tremelo jack on the back of the amp, allowing it to short out the circuit at the jack. He said he could do this, but he also suggested taking most of the tremelo circuit out of the signal path, using the switch to by-pass most of it.

He hooked up some wires to show me what he meant. We listened to the amp with the tremelo section on and the speed and intensity dials turned off; with the tremelo footswitch engaged, and with the circuit by-passed. The amp was noticeably clearer and had a more bell-like tone with the tremelo circuit completely bypassed. It was subtle, but I could certainly here it. Once again, simpler is better.

What Rick did was take the brown wire from the right tab of the intensity control and run it over to the old ground switch (now two-way). I’m not exactly sure of the connection back to the rest of the circuit, but you get the idea. With the switch off, the only part of the tremelo circuit engaged is the intensity pot.

By now it was a little after 10:00 and it was time to close everything up. As he was putting the chassis back in he noticed he had forgotten to change the third reverb capacitor. Perhaps that’s what was causing the leak. He said we’d get it next time.

After putting the chassis back in, Rick tested everything on the amp again, and it sounded great. Rick said that it could sound even better if I wanted to get some more work done on it. He thought changing the blue coupling caps might help, but pointed out that these are good sounding caps and until you try replacing them you won’t know if it will help or hurt. He also said when you do some big changes like today, you expose weaknesses in the rest of the circuitry. This turned out to be a prophetic statement.

We wrapped it up for the day. I thanked Rick for the amp lesson. He really is a nice guy, and must have a lot of patience to take the time to explain things to me for seven hours.

I carted the amp up the creaky basement stairs, said goodbye to the cat and rabbit, and went outside into the very dark country air. I had a great sounding amp and a head full of new ideas. Definitely my idea of a great way to spend my Sunday. I drove home, and of course plugged the amp in and played for a while until my wife wanted to go to bed.

Over the next week I played the amp everyday, and everyday it sounded fine, until last Saturday. I had been playing for about an hour when I noticed the amp had started to make crackling noises and was humming louder. I turned it off, waited a while and turned it back on. It was fine for a while, but it started again. By the way, these sounds remained even with the volume control turned off. I assumed a tube had gone bad, so I started replacing each one, one at a time including the rectifier tube. No luck. The same thing happened each time.

I decided to try to get something to blow to help the diagnosis. I plugged in a PRS and turned the guitar and amp up to 10. After a minute the sound quit and a puff of smoke came out the back of the amp. Success. I’ve seen a puff like that before on a Deluxe Reverb Reissue, and I suspected a screen grid resistor had blown.

This is obviously frustrating, but not surprising, and I’ll need to take it back to Rick’s and get another amp lesson to see what’s causing it. Perhaps I’ll get the blue coupling caps and the metal canister capacitor replaced while I’m there.

What a hobby! Up one day with the sweetest sounding amp, and down a week later. It’s frustrating, but eventually it was worth it.

Restoring the amp Part II: My 3 hour Tube Amp Lesson

I went back to Rick's, and sure enough a grid resistor had blown, taking out one of the Visseaux 6V6 tubes. Rick decided to put in 5 watt screen grid resistors (instead of the 1 watts) since he knew I’d be pushing the amp pretty hard.

The intermittent noise problem was due to a thermally sensitive cathode resistor on tube 3. Rick found it by heating up the amp and then carefully spraying each resistor and capacitor with a cooling aerosol spray (don't know the chemical.) You could clearly hear pretty loud noises from the cathode resistor when it was cooled down with the spray. After it was replaced, the problem was gone.

Next, Rick replaced the big canister cap, which tightened up the sound a lot, and then we tried the B+ phase inverter mod. It did as promised; increasing the volume of the amp, but it did change the tone a bit. I would categorize it as a cleaner sound with a lot more attack at the front of the note. We reversed the mod, retaining the browner/rounder tone.

Another thing we did was compare the Visseaux 6V6's to the Soktek 6V6EH's. Using a variac to control the power, and the 10 turn bias pot to have the bias on the tubes the same, we listened. The Visseauxs are beautiful with a tighter low end than the Sovtek's and noticeably more sustain. The Sovtek was pretty darn good though. I thought it sounded a bit harsher than the Visseauxs. The high end would reach feedback a little easier with the Sovtek's, which is an infinitely better tube than the Sovtek 6V6GT's.

Still in pursuit of more clean power with more sustain, I agreed to try replacing all the blue caps with the Sprague Orange Drop caps. This did change the tone of the amp, but not for the worse, and boy-oh-boy did it ever increase the sustain. If I learned anything during these 10 hours at the bench, it is that sustain is all about quality components, especially high quality capacitors. While Rick was at it he replaced the reverb capacitor he missed during Part I.

And now for the coup d’etat. We replaced the Phillips NOS 5U4GB tube with a NOS Mullard 5AR4/GZ34. This is not typically recommended, but it was an experiment that turned out great. Please remember that putting in the 5AR4/GZ34 rectifier tube increases the plate voltage on the 6V6 tubes beyond what is recommended for this amp. But knowing that the new Sovtek 6V6EH’s are designed to take 500 volts on the plates, we decided to give it a try.

After re-biasing the amp, there were 435 volts on the plates, and the Sovtek6V6EH’s sounded even better than they had before at a lower plate voltage. I decided to keep them in the amp and save the Visseaux 6V6’s for another application.

The very surprising thing about putting in this rectifier was the obvious improvement in the sustain and harmonic complexity of the amp. Don’t let anyone tell you that rectifier tubes have no effect on the sound. They do! I can hit a low E note on my Strat and it will almost sustain forever (without distortion) and the note swells and recedes and swells again in a most glorious fashion due to the sag effect, which provides compression. Of course all of the tweaking of the amp up to this point certainly helped, but I was amazed at what this Mullard 5AR4/GZ34 did for the amp. No wonder they are hard to get and sell for so much.

Now I've got a Princeton Reverb that just drips with tone. I’ve gigged with it three times with a loudish drummer, 2nd guitarist, and bass, and it hangs in there just fine. I took a solo on a Jefferson Starship tune and turned the reverb up just before playing the solo. On the first note, people in the room turned, looked at the amp, and smiled. Impressive. For low volume playing at my house it sounds glorious too. 12/24/99