Reverb and Driver tube: 12AT7 or 12AX7?

[6267]

ecl82 schemo

I've attached the approximate schematic of what I did to use the self-split (cathode-coupled) ecl82 reverb driver. You must *not* bypass the cathodes, since the audio signal is made push-pull by 'sharing' it through the cathodes. The resistor values and cap values warrant experimentation! I believe everything I did is there, and this circuit gives the benefit of leaving one free triode for other purposes. It would thus work as a tube tube standalone reverb, with the 'extra' triode driving the input. I think by decreasing the screen resistor output power should increase. The 1 meg grid leak may be high.

[Steve Conner]

I like the self-split reverb driver too. I've used a 12AU7 self-split driving a small "Universal Output Transformer" that I had lying around, and it seemed to work better than paralleling the two triodes and driving the same transformer SE.

[6267]

There are some errors in the schematic I posted, there should be a 100k grid leak for the gain stage after the tank, and the anode resistor was 100k, not 220k...I ended up removing the cathode bypass cap as well, I had way too much gain in the reverb return.

[martin manning]

The 6G16 brown Vibroverb used parallel 12AX7 triodes to drive its reverb. Looking at the schematic I see 440V on the reverb transformer, 430V on the plates of the 12AX7, and 3.3V on the cathodes via a 1K cathode resistor.

So, that means 1.65mA through each side for 710 mW plate dissipation, which is okay, but it looks like 1.65 mA and 430V is off the map for that tube. What was Leo thinking?

MPM

[Steve Conner]

Probably something along the lines of "How do I use up this huge heap of 12AX7s in the corner of the factory"

Another example of off-the-map tube application is Tim De Paravicini's V20 power amp, that used 20 12AX7s as power tubes. http://www.stereophile.com/integratedamps/1099ear/

[mac1amps]

I've converted a number of Fender reverb drives (in my own Fenders, as well as a number of my clients) from 12AT7's, to EL84's, primarily due to the fact that since an EL84 is a power tube, it will last ALOT longer than a 12AT7 ever would. They seemed to have a wider dynamic frequency response as well. I've also done it to a few Dr. Z's, which had [IMO] a wierd network that used a paralleled 12AX7 in the POST spring pick-up stage (????), but only half of a 12AX7 for the drive. My client was much happier with the reverb operation after I changed the drive to the EL84, and just used half of the 12AX7 in the pick-up stage.

[GainFreak]

I've converted a number of Fender reverb drives (in my own Fenders, as well as a number of my clients) from 12AT7's, to EL84's

Didi you use the Fender stock reverb OT or any other type?

[mac1amps]

All of the EL84 conversions were done using the stock Fender 022921 tranny, although with that said, I'm hoping to soon get around to re-building my re-issue "stand alone" Reverb head, at which time I was going to experiment using some alternate 5-10 watt single ended xformers, 'just to see what kind of difference it will make (if any).

[Sir Cuitous]

mac1 - what value of resistor do you use for the EL84 cathode?

[mac1amps]

It varies a bit, depending on the model, but it's usually in the "ball park" of about 820 Ohms (I like using a 10 watt). I also use a 470 Ohm (1-2 Watt), tapping off the preamp supply for the screen grid.

[mooreamps]

It varies a bit, depending on the model, but it's usually in the "ball park" of about 820 Ohms (I like using a 10 watt). I also use a 470 Ohm (1-2 Watt), tapping off the preamp supply for the screen grid.

Sounds like you are on the right track. congrats .....

-g

[Sir Cuitous]

mac1: good info, thanks.

[GainFreak]

I didn't want to start a new similar topic that's why I got this one up.

I was always curious how that Mesa reverb circuit will sound and if they are using their own transformer.
Recently I built the circuit but I used a 022921 I had by hand and I have to tell you the reverb sounds very acceptable despite that theoretically the OT should have higher impedance. I know it's a matter of taste but for my needs the reverb sounds OK. I tried a 12AT7 in the same circuit and sounded weak at the recovery side which is to be expected.
I was somewhat confused by the high voltage at 12AX7 plate (I have around 400V there) but after I did some measurements it turns out that the driver part draws 2mA of current. At 400V this makes 0.8Watts which is well within the 1W 12AX7 anode capability so despite the higher voltage the tube is not overdissipating.
Also I noticed that in the schematic Mesa's reverb driver OT leads color code is the same as the 022921. Is this a coincidence? I don't know because I've never measured a Mesa reverb OT but I doubt it.
Because I'm not very good in theory (I like more the soldering iron) I wanted to ask how the uncoupled 22k resistor coming from the screens and going to the reverb OT affects 12AX7's load impedance. Does it go in parallel with the anode, the transformer, with both or what? And what is the resulting impedance and what should be the tube life expectancy if running it with a 022921?
I would appreciate if someone explains that.

[Gingertube]

Now if ONLY we could get a tube manufacturer to re-release the ECL86 (6GW8), a 12AX7 spec triode plus a slightly down rated EL84 ish pentode (actually its more like an EL80 or 6M5).

I've also seen an ECL84 (6DX8) used for reverb drive/recovery.

I guess we are stuck with ECL82 for a triode/pentode tube from modern production.

For those who like NOS tubes there are a S..t Load of 6GV8 (ECL85) around cheap. They might be worth looking at.

My own home brew guitar amp uses push pull 12AT7, the push pull tanny is a Hammond and the circuit is "borrowed" from Kevin O'Connor. It gives good reverb depth (what guys like to label as "dwell").

Also I spent quite some money and time experimenting with different tanks. For Rock and Blues I settled on the 6 spring, long delay Accutronics. The long delay was better than the medium delay although I've heard that "shreader" players often prefer the medium delay.

In both long and medium delay, the 6 spring tanks were MUCH better than the 4 spring units.
I only looked at the 17"? size tanks - Anyone with a Marshall or a Mesa will be able to tell you that those little tanks (the 9 or 10 inch ones) are a waste of time if you want decent reverb.

Cheers,
Ian

[GainFreak]

I finally had the chance to measure a Mesa reverb transformer and my suspicions are now confirmed. It's a Fender type of reverb OT.
The original Fender transformer has a transformation ratio of ~61 (220V/3.5V), this one has ~56 (19V/0.34V).
Primaries DCR and inductance also match within 5%.
I'll be unwinding the cheap New Sensor (Taiwanese) version and will let you know the winding data if you're interested.