12AT7 reverb tube for Deluxe AB763 circuit

[mikepukmel]

Hi All
Working on adding reverb to a deluxe build that didn't have it. I noticed that the 12AT7 first reverb tube (the send tube) plate is powered from the first filter cap, right after the choke (and before any dropping resistor), marked "B" on the AB763 schematic.
On the old schematics found online, the voltage at this point is marked 410v, but I know in my amp its much higher, probably over 430 or 440v.
After getting the additional tubes for the reverb, I noticed on the data sheet for Electro Harmonix 12AT7 max plate voltage 330v. <big frown> Its not even in the ballpark. So I started looking around at various tubes online and haven't found one that has a max plate that high.
Then I noticed that every other tube in this amp, and most any other amp schematic I could find, have a 100k resistor just before the plate feed, but in this schematic, there is no resistor between point B on the power supply, and the transformer primary lead.

Can someone recommend a new tube for this spot in the amp? With new production tubes, do you guys put additional components in to drop the voltage to this tube on a replacement?

As another aside, Ive seen 'complete tube set', some have JJ's in this spot, and their data sheet says 300v max!

Thanks!
Mike

[Leo_Gnardo]

If you're all that worried, add a resistor 10K to 20K, say 3 to 5 watts, between the B+ node and reverb transformer primary. I'm not sure that will put the plate at 300V but it ought to make life a little easier for your 12AT7. Also consider the millions of old Fender amps out there with tube-driven reverbs, they're not failing left and right. On occasion I see a drive tube that's failed, or gone noisy, or showed its disapproval by smoking its cathode resistor. Also I've had a rash of output transformers gone noisy in old (1964 thru 1985) Fender amps, maybe 3 or 4 a year out of hundreds of Fender amps that hit my workbench.

[Randall]

Shouldn't we be worried about the extra current draw on the PT from adding two tubes?

[Enzo]

Fender has been using the 12AT7 in this same circuit for 60 years or more. The tube data sheet numbers are for polite table radio use, We routinely run 6V6s in amps at 100v over spec. We don't need to change that at this point. You don't need a different tube.

When you are expecting 410v, 430v is not "much higher". That is only about 5% high. Look in the upper right corner for the schematic notes, see where it says all voltages within +/-20%?

In a tube, the grid voltage controls the current through a tube. Your typical 12AX7 triode has a 100k plate resistor to give that current a way to make a voltage change. The 100k resistor is the load for that plate. In the case of the transformer, the current flowing through the tube also flows through the transformer primary. The transformer is the load, so it doesn't need a resistor.

If we followed the tube data sheets, we wouldn;t have many tube amps around. This circuit has been effective and reliable since your grandfather owned one, no need to change it now.

[mikepukmel]

Thanks everyone!

Leo, thanks, can do, Is that what you guys do when you re-tube an old amp? Do you find that replacement 12AT7's in this spot in the circuit burn up faster than the original tubes did?

Randall, thanks for the heads up, the parts should be for a Deluxe Reverb: the transformers, and chassis for tremolo and reverb, but I was stupid and didn't order the parts for reverb when I first built the kit. Mea culpa. Ended up with a 2 channel 1 channel amp, that should have had reverb wired in when at first. Anyway, the Hammond transformer 291bx is for a fender deluxe reverb. It will still be one tube short since I didn't put tremolo in.

Enzo, thanks, I always forget about tolerances! maybe its to be expected, I just don't know what to expect. I didn't word it correctly, but I was worried that if the schematic had 410 in this spot and my amp is even higher, even if its in the 20% band, that's farther from the 330 max rating. And the more tubes I look at the more I find the same, they're all 300 to 330 rated max.

How DID these amps work like that? Did these tubes in this spot in the circuit burn out faster, or were older tubes made to handler the voltages better? Or am I worrying about nothing.

[mikepukmel]

Enzo, words to live by:

"If we followed the tube data sheets, we wouldn;t have many tube amps around. This circuit has been effective and reliable since your grandfather owned one, no need to change it now. "

All great news, so I can plug in the tubes, and fire it up! Well, ok not exactly "fire" it up Heh. Energize the tubes and annoy the neighbors with my Jeff Beck covers.

[Randall]

" Did these tubes in this spot in the circuit burn out faster, or were older tubes made to handler the voltages better? Or am I worrying about nothing."

no, yes, and yes.

[Chuck H]

Honestly, just because it's pretty easy I would just branch a 15k/5W resistor and add a 22uf/450V filter to the end of the branch. That will probably get you a little below the Fender spec, closer to the tube spec and WAY below the 120 to 125VAC into a vintage amp figure. Just because Leo did it and it works doesn't mean you have to abuse the tubes also (at least not as badly). Two components and making a place to put them. Easy peazy.

Something to consider about the Fender amps is that the later cathode R spec for the reverb driver seemed to increase in value. An easy fix to reduce dissipation, but it does nothing about the high plate voltage. Proving that the spec has A LOT of leeway. So, if you do reduce plate voltage for the driver you'll probably want to use an earlier than later cathode resistor value spec (or you could check the actual current and adjust it for your own purposes).

JM2C

[g1]

Here's a different model Fender with the same reverb circuit, 12AT7 plate voltage shown as 440V. That would be even higher with modern line voltages.
So you don't need to worry.

[Leo_Gnardo]

Something to consider about the Fender amps is that the later cathode R spec for the reverb driver seemed to increase in value. An easy fix to reduce dissipation, but it does nothing about the high plate voltage.

Good observation! When I replace roasted Rk's on reverb drive circuits, I generally use a 2200 ohm resistor, 1 to 3 watts depending what I have on hand. I've seen Fender use 2200 half watts down to 470 ohms. This tube does take a beating, and IMHO there's more than enough drive signal to spank the tank. A little reduction here isn't such a bad thing. I like your proposal for a branch off the main B+ supply too. There also the values can be altered to suit whatever you have on hand or can get cheap. For instance 10K, 10 uF, close enough for rock n roll. Especially if you're entertaining your neighbors with Jeff Beck licks. What must be going thru their minds, "are we being visited by flying saucers??? Martha, go have a look out the window!"

[Enzo]

It isn't plate voltage that stresses a tube like this. Excess heater to cathode voltage can be an issue. Not in this circuit, but in a cathode follower say. The other might be cathode current.

The plate voltage is not hurting the tube. The reason they used a 12AT7 and the reason they also paralleled the two triodes is they wanted sufficient current drive through the transformer. But looking at the schematic, I see 8.7v across 2200 ohms cathode resistor. I get about 4ma cathode current from that. About 30 milliwatts resistor dissipation.

I don't see any threat to this tube here. And indeed, I have replaced a reverb drive tube now and then, but no more than any other tube in an amp. I don't see Fender amps eating reverb tubes.

[Chuck H]

Agree! The current is usually low enough that the tube isn't stressing for dissipation.

But look at that circuit g1 posted!?! 440Vp and a 470R cathode resistor. With 6V on the cathode that maths out to 5.6 watts idle current. I know I've seen some Fender reverb drivers that seemed actually physically hot. This must be an example.

[Enzo]

REally? I get V sq over R for power. (6x6)/470. 36/470 = 77mw? A half watt resistor ought to be fine.


Oh wait, you mean tube dissipation. Even so, I don;t see amps eating tubes.

[mikepukmel]

Thanks everyone. I owe you all. Going to print this thread and put it in my binder!

Think i got it: the original designed worked. many amps ran decades with no problems. (I've had a few). Plate dissipation is the core issue, and even with high plate V, not a problems since not a lot of I through this tube. But, no reason to beat the tube, so might consider putting a few more components in to drop the plate voltage. OK I get it, Fender didn't do it since they were a production shop, fewer components, cheaper, easier to manufacture, and ... the amps had good longevity.

thanks.

[mikepukmel]

Here's a different model Fender with the same reverb circuit, 12AT7 plate voltage shown as 440V. That would be even higher with modern line voltages.
So you don't need to worry.

There is a 560pf 3kv cap across the primary of the reverb driver transformer. Is this an oscillation killer cap?