Ampeg Reverberocket R-12-R DIY project

[andrewRneumann]

How does V3A get away with running continuously at 2.6W (service manual volts)?

 

[Lowerleftcoast]

Doesn't introducing AC into a filter cap induce ripple?

Technically ripple is from rectified AC. It is strong pulses derived from the mains (60Hz) frequency. Here in the US, full wave rectified, the pulse will be 120Hz (double the 60Hz) but there are some harmonics at multiples of 120Hz. The filter caps and dropping resistors make low pass filters which let all frequencies above around 1Hz, (yes, that is like one wave per second), to ground. The ideal is to have pure DC power with all AC content filtered away.

Signal AC is not strong and it is many frequencies. There may be some anomalies like Andrew mentioned but signal AC makes it past the plate resistors all the time. A little more through these caps will be filtered away.

First, because it is worrying me, I've done calculations for the HT current draw

I am concerned about the 90mA PT as well. The real test will be how hot the PT is under load. Since there is some good info about a 90mA PT being original to this amp along with the knowledge the Princeton had a PT rated at 70mA... I am optimistic this PT will pass muster. Time will tell.

Sonny Reverb has reported an issue with microphonic noise on V2, the pre-and post-reverb intermediate amplification stages

Well the larger bottle is known to be more microphonic.
I realize the 6SN7 has far less gain than a 12AX7, but it is reverb driver. Think of it as single ended amplifier. Instead of driving a speaker this transducer wiggles springs. In the Princeton it is like diming one of Rob's micro amplifiers 100% of the time. Lead dress and maybe even backing off some gain is likely very important just like in a Princeton Reverb.

Just how hard is that tube being pushed? Something to think about.

 

[chas.wahl]

V2 has the signal amplification triodes that, in the chain of things, precede the reverb driver (V2a > V3a) and follow the reverb recovery (V3b > V2b, and thence to the PI. Both V2 and V3 are 6SN7s. Unless you're saying that the reverb driver is affecting the tube (half, V2a) preceding it, in some way, I don't understand how V2 being microphonic is because of the reverb.

What had occurred to me (in my untutored musings) about microphony problems with the V2 tube: isn't one of the effects of having the snubber cap supposed to be to prevent oscillation? Isn't microphony a result of uncontrolled oscillation? Well, the V2b section has such a snubber, but the V2a section (the 1st section following the input triode V1a) doesn't have one. Could that be the culprit?

One thing that's different about this reverb circuit, compared with, say, the reverb on a Princeton Reverb (AA764 or AA1164) is that Ampeg's switch closes to shunt the output signal from the reverb driver to ground, while the Princeton's footswitch shunts the input of the recovery triode to ground.

How does V3A get away with running continuously at 2.6W (service manual volts)?

Maybe I'm don't have the electronics smarts to see what the problem is there. The datasheet for the 6SN7 has it rated at 5 W/section, and 7.5 Wcathode for combined for both sections. I've pored over a lot of information about this amp, both factual and anecdotal, and I don't remember folks complaining about it burning out preamp tubes.

To me the weird one is the V1b trem section, that runs at 310 V when trem is off. Now that exceeds the "max. Volts" in the 6SL7 datasheet, if only by 3%.

But, I'm pretty dumb about those graphs at the end of datasheets, really must work harder at that and educate myself (again) about them.

I am concerned about the 90mA PT as well. The real test will be how hot the PT is under load. Since there is some good info about a 90mA PT being original to this amp along with the knowledge the Princeton had a PT rated at 70mA... I am optimistic this PT will pass muster.

My hunch is that, like Fender with the Tweed Deluxe and Princeton Reverb, Ampeg was expecting some PT "sag" in voltage as a trade-off to suck more current out of it -- a cost consideration. One of the benefits of having more primary voltage now is that the transformer may well put out the required current at a somewhat higher voltage than it used to with 117 V supply, but as you say, it may get hotter. I'm not the sort of person who's going to overtax an amp, though -- I play on the "polite" side, under "bedroom in a big apartment building" conditions where I pretty much have to be.

I guess what I was really asking is: how realistic/reliable are my attempts to quantify the current demand?

 

[Lowerleftcoast]

I don't understand how V2 being microphonic is because of the reverb.

You got me there. A tube will not be microphonic because of reverb.

I am reading more into his statement than was actually there. I am associating the reverb driver squeal a Princeton Reverb demonstrates with his *microphonic* comment.

 

[andrewRneumann]

Maybe I'm don't have the electronics smarts to see what the problem is there. The datasheet for the 6SN7 has it rated at 5 W/section, and 7.5 Wcathode for combined for both sections. I've pored over a lot of information about this amp, both factual and anecdotal, and I don't remember folks complaining about it burning out preamp tubes.

Oh ok. The datasheet I was looking says 2.5W per section. I guess it varies.

 

[chas.wahl]

Well the larger bottle is known to be more microphonic.
I realize the 6SN7 has far less gain than a 12AX7, but it is reverb driver. Think of it as single ended amplifier. Instead of driving a speaker this transducer wiggles springs. In the Princeton it is like diming one of Rob's micro amplifiers 100% of the time. Lead dress and maybe even backing off some gain is likely very important just like in a Princeton Reverb.

Just how hard is that tube being pushed? Something to think about.

Somewhere in my peregrinations today I saw the word "SELECT"; where was it? Oh, on Peegoo's "prototype" R-12-R schematic from 1961:

Somebody was having some kind of problem, seems to me. Note that the 0.0068 uF cap is missing from the Dimension pot on V3b, and the "snubber" cap on V3a is only 0.01 uF, rather than 0.02 as on the 1962-09 schematic. Also note that the footswitch for the reverb is shown on the recovery side, but someone has x-ed that out, and drawn connection to the driver side (Ampeg? Amp tech?) as it is on the later schematic. Plus: the plate and cathode resistors on V3a are 5.6k and 1.5k, while on the 1962-09 schematic those have been changed to 2.2k and 330 Ω.

 

[JohnnyCrash]

I’m guessing they kept improving on the circuit as customer feedback trickled back in? Or as different replacement tubes handled (or couldn’t handle) things differently, designers tightened up the troublesome bits of the circuit. Fender did a lot of tweaking like that to a given circuit. I have no idea what other audio/radio manufacturers were doing. But maybe those were improvements?

I’m too medicated to look through the schematic though, so my blind guess could totally be off.

 

[chas.wahl]

I'd like to ask a question in general about the signal path. With reverb and trem on, it's like this (I think):
V1a input > V2a > V3a reverb driver > reverb tank > V3b reverb recovery > V2b > V4 PI. Trem effect is inserted post-PI after the coupling caps to V5 and V6. Unless there's a split in the "dry" and "wet" signal that I've missed.

I can't seem to figure out how the signal chain works when reverb is off. There's that 22k resistor between the cathodes of V2a and V2b; is that a separate path for the dry signal continuously? Or an alternate path if the reverb is incommunicado? Or does signal somehow travel through the driver and tank and recovery when reverb driver's signal is grounded by the footswitch?

The back-and-forth in this amp puts me in mind of Dr Seuss (Oh, Say Can You Say): "The words in this book are all phooey. When you say them, your lips will make slips and back flips and your tongue may end up in Saint Looey!"

 

[Lowerleftcoast]

There's that 22k resistor between the cathodes of V2a and V2b; is that a separate path for the dry signal continuously?

Yes.

 

[NTC]

Isn't microphony a result of uncontrolled oscillation?

No. It is mechanical vibration (sound) coupling into the tube and making the parts inside vibrate audibly. If the sound stops, the microphonic sound stops in many cases - unless the tube is really bad. If you filter out the frequencies that cause microphonics, you are very likely to filter out signal you want as well.

 

[chas.wahl]

OK, so snubbers don't have anything to do with microphonic tubes. And microphonic behavior has only to do with physical exposure to sound waves or other vibration? So the only fix is to isolate tubes that tend to be microphonic is physical isolation from such vibrations?

Are snubber caps intended (according to something found by googling) "to reduce the impedance rise at high frequencies due to the parasitic inductance of the main electrolytic capacitors"? I also read something about them helping to reduce the current load on the plate resistor and plate due to higher frequencies.

Getting back to component specifications: is there a reason to source snubber caps, especially those that don't seem to be available as mica caps (just the 500pF ones) as a higher voltage-rated cap than the Ampeg 400 V spec? That will make them larger, especially if sourcing film & foil rather than metallized polyester or polypropylene (which tend to be more compact). Ampeg went with all paper, for both coupling and snubber caps, except for the oddball in-between 0.0068 uF cap, which was ceramic -- I'm inclined to avoid ceramic caps.

 

[andrewRneumann]

OK, so snubbers don't have anything to do with microphonic tubes. And microphonic behavior has only to do with physical exposure to sound waves or other vibration? So the only fix is to isolate tubes that tend to be microphonic is physical isolation from such vibrations?

That‘s pretty much it. Using flexible strand wire and suspending the tube on a trapeze net seems to be the ultimate solution. Maybe a tempurpedic mattress would help.

Microphony can result in oscillation—just like a microphone can feedback. It can ring at a resonant frequency or it can go full blown feedback at resonance depending on the damping.

Are snubber caps intended (according to something found by googling) "to reduce the impedance rise at high frequencies due to the parasitic inductance of the main electrolytic capacitors"?

This is true if the caps are installed in parallel to the main electrolytics. When we are installing the cap from the plate of the tube to AC ground, we are just making an RC filter. I don’t think I’d call your situation a snubber. Come to think of it, I hear the word “snubber” applied to all kinds of caps that it has begun to lose its meaning. Not sure I even know the official meaning of “snubber”.

I also read something about them helping to reduce the current load on the plate resistor and plate due to higher frequencies.

It does bypass the plate resistor, but it should increase the AC current through the tube because there is less AC impedance at those frequencies. Keep in mind that the voltage swing is reduced, so there is not necessarily any increase in the power output of the tube at those frequencies. More current swing doesn’t mean more power if voltage swing drops simultaneously. I wouldn’t worry about it.

 

[chas.wahl]

In trying to understand how the PI on this amp works, I found this online article by Richard Kuehnel:

Ampeg Plate Follower Phase Inverter

www.ampbooks.com

Though the specific model he discusses is the M-15, with a couple minor exceptions, I think it's almost identical to the one in this amp. I made a PDF leaving out the interrupting ads, so if anyone would like a copy of that, just PM or email me with an address and I'll send it.

 

[Lowerleftcoast]

Not sure I even know the official meaning of “snubber”.

I agree it has been misused, and I am guilty of misusing the word. I think it is more correct to use the word *snubber* to describe a cap across a switch.

I believe the correct term for these caps on the Reverberocket would be shunt. It shunts high frequencies.

Imo, the reason for these caps on this amp is to reduce audible highs.

I wouldn't be concerned with the AC current at these high frequencies.

The voltage rating just needs to be larger than the expected voltage the cap will encounter. These caps are not in the signal path. A ceramic disc cap can perform this function. I choose to avoid ceramic disc caps mainly because the accuracy over time (drift) is questionable.

 

[Pharmerdavid]

Yes, you're correct.

Based on what I've learned about 'em, this is a very early Reverberocket. Another clue is on on the tube chart (the handwritten "R" after the model), because they did make R-12 Rockets before they made Reverberockets. They used a Rocket label in this one and wrote an R there. This info I got from Ken Fischer (RIP), who worked at Ampeg when these were being built.

Being that the Reverberocket was the first amplifier built with onboard reverb, it makes this particular amp pretty kewl.

I'll take a peek under her skirts tomorrow when I get some time, and record some voltages.

Here are pics of the PT and OT. It does have a PT-106 installed.

'926' on the PT is the EIA code for the manufacturer--Todd Electric Company of Yonkers, NY. '129' is the production date: 1961, week 29.

'682' on the OT is the EIA code for Electrical Windings Incorporated of Chicago, IL. '112' means the 12th week of 1961.

More info as I get time... Cheers!

 

[Pharmerdavid]

Not to be contentious, but in the interest of accuracy: Ampeg did claim the Reverberocket was the first combo amp with onboard reverb, but Gibson did it in 1961 too, and Premier did it in 1959 - two years earlier, so Ampeg fibbed (Gibson never made any claims ). Also, Ken Fischer worked at Ampeg later in the 1960's, not when these early Reverberockets were being made, although he did mention the 6V6 Reverberocket being an excellent amp with great overdrive, the reason Everette Hull had them switch to 7591A power tubes for less distortion. I own a 1961 model, looks just like the one in pictures posted earlier in this thread. I also own a 1963 model R-12-RB with 7591 power tubes and SS rectifier. When I get my R-12-R (6V6) model back from my tech, I may try a SS rectifier in place of the tube rectifier, to tighten the lows and reduce squishiness, although I may like it squishy - we'll see. I always change the speakers, even though the old Jensen's sound great, the replacement speakers with larger voice coil and magnet sound much better. The 1961 reverberocket came with a P12S speaker, which is a tiny magnet, and it doesn't do the amp justice. I have a Weber 12A150 50 watt in it now, and looking forward to getting it back from the tech!