ampage Tube Amps / Music Electronics	For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

Small sig. pentodes - EF86 (Long)

:

9/13/2000 5:08 PM

Small sig. pentodes - EF86 (Long)

Hi guys,

I know there was some talk here awhile back about these tubes, as well as people's homebuilding experiences with them.

I've been monkeying around during the last couple days, trying out some basic preamp topologies, with very mixed results.

Granted, I'm no rocket scientist, but I've been cobbling the EF86 on the front end of a number of circuits so I can hear what they sound like. Obviously there is the potential for incredible gain compared to a 12AX7, so I was expecting some chaos (zero design sense on my part.)

I used the topology from the Svetlana website as my gain stage http://www.svetlana.com/graphics/TB/no.23fig1.jpg

In my first attempt to use this circuit I added it to the front end of a 5E3 knockoff (in lieu of the 12AY7) Sound wise, it was pretty over-the-top, lots of distortion. When I took voltage measurements I was surprised to see that the EF86 screen voltage was almost twice as high as the plate!!! I ended up increasing the screen grid resistor which got the voltages more in line, but the noise level went up and the tone was horrible.

After double checking my wiring and finding everything was hooked up right, I soldered in the EF86 circuit for the first stage in a SE Tweed Princeton knockoff (5F2A) I essentially got the same results.

Being in hardcore tinker mode I hammered together a trem-less AC4 today and the results were much better. With the high screen grid resistor (5M) the plate voltage was about twice as high as the screen.

Would having too much gain (i.e. following the EF86 with the 12AX7 stages as I did in the "Fender" circuits have any influence on my wacky voltages? Is it just a question of bias?

My apologies for the length of this (and overall denseness on my part), but if anyone has any tips I'd greatly appreciate it!

Jim

9/14/2000 12:05 AM

I think you could tweek a few things, that's a tremendous amount of gain for a guitar input circuit to have. I doubt if the voltages you are seeing are meaningful. Make sure the control grid is grounded by by some means of a load resistor. You could also lower the bias by increasing the cathode resistor if you like.

One way you can tame the EF86 is by lowering the screen voltage. Rather than a 1 meg to the Vp, try using a divider network to get say 1/2 Vp to the screen. Try adding a 1 meg to ground at the pin 1 and 1 meg screen supply resistor.

I'd also stick with metal films on the cathode and plate resistors to help control noise. Perhaps a ceramic socket too. Some kind of grid stopper on the input might be advisable as well.

9/14/2000 2:56 PM

Hi Nuke,

Thanks for the info!

I should have mentioned that I have a 68K input resistor with the 1M to ground. I ended up using a .022 coupling cap to the next stage.

I know that pentodes are definitely another animal compared to triodes (that pesky screen grid 'n' all)...I guess what flipped me out the most was that my voltages were pretty wacky for what I thought was just going to be a cut and paste circuit (the Svet one).

For some fun and frivolity I'm going to use pots for the plate and screen grid resistor and a decade box for the cathode and just screw around with the resistance coupled amplifier data in a couple tube manuals.

Thanks again,

Jim

9/14/2000 5:00 PM

The makeup of the circuit following the EF86 really has nothing to do with the voltages you are seeing on the plate and screen of the EF86. In my EF86 homebrew, I have a 300V B+ rail feeding the EF86's 220K plate resistor and 1M screen resistor. My screen voltage is higher than the plate; if I remember correctly, I'm getting 80V at the plate and 115V at the screen, but I could be off a bit. This is normal for the EF86. You can't just drop this tube in place of 1/2 of a 12ax7 in a given amp design and expect much. The EF86 will produce over 3 times as much voltage gain as 1/2 12ax7, which your amps are not designed for. The result, as would be expected, is an overloading of the grids with signal at every point after the EF86 in the amp. Use a 220K plate resistor and a 1M screen resistor with a bypassed 2.2K cathode resistor on the EF86 and it should be working perfectly. You just need the proper stages after the EF86 for a design to work. You might want to look at a schematic for a Matchless DC30, which has an EF86 as the input tube. On my homebrew, I have the EF86 setup as described above driving a 6SL7 in a long-tailed pair PI configuration. Good luck.

Page 1 of 1