EF86 needs more drive?

[JoeK]

I'm trying to decide on a couple of preamps for a Vox 30watt build. Had decided to have a rocket/top boost pre on one side and an EF86 pre on the other channel. I was hoping the EF86 channel might be giving some small pentode overdrive before it hits the power amp but it doesn't seem to get close. I've borrowed a few pieces and after trying different things it looks like this.




maybe I've forgotten how to attach pics but here's the direct link (I hope)
http://i188.photobucket.com/albums/z.../Ef86Pre-1.jpg

So roughly I've got the paralleled triode from a "Clubman" and put it in front of Tubeswell's tonestack and EF86 morph setup. I initially had a single triode with traditional 100K/1.5K/25uf before the tonestack with the morph setup. I found without a master volume post EF86 the noise from the pentode was too intrusive, it's just tube noise not heaters or ground etc. I had an old Telefunken EF86 then got a Mullard EF86 plus a Phillips (german) EF86 which were a big improvement. The EF86 channel just has so much output compared to the Top Boost channel I thought I would drop some gain on the EF86 (shorting the input grid of the EF86 did not affect the noise generated by the EF86) Sorry for the long story but thought it might explain some things?
So I've put a humbucker into the HI input and opened all the gain up but I can't hear this EF86 breaking up as I hoped it might. Of course if I open the Master volume the Vox AC15 power amp setup I've got goes into overdrive but I was hoping the preamp might do that by itself. Any suggestions? wrong preamp? Thanks in Advance Joe.

[tubeswell]

Hi Joe

For some reason your attachment link didn't work when I tried it just now.

[JoeK]

Well I did a bit more testing with this and tried to see how much I could get out of the EF86. I had not used Tubeswell's "raw" pot for lack of room on my test chassis. It is a 1meg pot on the bottom of the tonestack to lift the earth reference and give a boost to the circuit, so I hooked up that pot temporary and removed the output tubes as I wasn't sure how much signal they'd get.
My voltages with a 200mv 1Khz input, Tonestack centred, EF86 pentode mode, were post parallel triode 350mv EF86 input at grid 60mv EF86 output 3.2V. Then with the tonestack lifted I had EF86 input 245mv EF86 output 13V which was impressive but the scope showed the output signal of the EF86 as being quite clean (to my eyes) The actual sound of a guitar with the input to EF86 full gain and post EF86 volume on 1 sounded rather thick/dark so I guess I got closer to getting an EF86 into distortion but it seems it may never sound like I've got an EL84 single end grind happening?

[loudthud]

What's going on with the first triode? 10K plate resistor and 2.2K cathode resistor? Is that right?

[JoeK]

What's going on with the first triode? 10K plate resistor and 2.2K cathode resistor? Is that right?

That is correct. I was trying to find a similar topology to the circuit and found the Matchless Clubman which has a parallel 12Ax7- tonestack- 6SH7. Tubeswell's circuit has a parallel 12AY7 with 100K on the plate and 820/22uf on the cathode. I will probably try tubeswell's para input stage with a 12AX7 or 12AT7 (as I don't have a 12AY7)

[tubeswell]

Hi Joe, try bypassing the EF86 Rk with 10uF. Also your schematic doesn't show any bypassing of the parallel input stage. You could try 22uF there.

(and if you need yet more gain, then try upping the plate resistor, screen resistor, and cathode resistor progressively to get closer to Voxish values).

[es345]

Hi Joe,

please keep in mind that with 6V or 13V at the output of the EF86 you are far away from overdrive.
Looking to the spec of the EF86:
with 250V of Ub, 100K Ra, 1K bypassed Rk you get 5% distortion only while having an output of 50V eff.

[imaradiostar]

I have different, possibly silly thoughts. TMB control between the two is killing all the gain you made with the 12ax7. I think this is a problem on the Clubman as well but not as severely.

Put the tone stack after the pentode. Bypass the pentode's cathode with a large cap as mentioned previously. Use a small cap (1nF or so) into a 250k "gain" pot between the triode and pentode. That'll hit the pentode hard enough to get it to do something interesting.

Let us know what you find!

jamie

[JoeK]

OK I put the first gain stage back to maximum gain to test the voltages and with a parallel 12AX7 and 100K plate resistor (120VDC on plate) Cathode has 25V/22uf-820ohm and 1VDC. Same 200ma input I get 6VAC at output of gain stage. TMB tonestack all centred and lift pot not engaged gives me 1VAC at EF86 input with 50VAC output. With the lift pot engaged I get 4.25VAC input at EF86 and 100VAC output to Phase inverter tube so she's an angry beast to be sure.
The scope at the EF86 output is showing the bottom of the sine wave starting to bend in on the trailing edge with the tone stack in place. When the tone stack is lifted and the voltages are maxed I'm seeing a huge flat spot top and bottom of the sine wave so I think I'm getting closer to the EF86 distortion I was after? Listening tests are a bit difficult to judge as I can't tell (even with the master volume very low) if it's the phase inverter distortion or EF86 distortion I'm hearing. I guess the next step could be a voltage divider to drop some signal after the EF86 and before the MV or is that just duplication and ineffective? Thanks again Joe.

[tubeswell]

sounds more like it. Is the EF86 cathode resistor bypassed?

[imaradiostar]

I like the idea of distorting the midrange strongly and having the TMB tone stack with a the mid scoop after the pentode. All the tone shaping in front of the pentode (mostly bass rolloff) helps the pentode distort a certain way the the tone stack "cleans up the mess" so to speak. Your mileage may vary, of course.

I second what Tubeswell said- you need a cathode cap on the pentode if you don't have one already.

jamie

[JoeK]

I haven't put the cathode cap back on the EF86 yet. I'll try that now.

[JoeK]

Well the cathode cap gave me a bit more gain and I measured 108VDC on the PI input. So if I open up the pre EF86 volume and have the guitar volume on 10 I can hear the distortion happening on the preamp. The post EF86 vol has to be on about 1 or the amount of signal coming out of the EF86 makes the amp sound like it's on 10 when it's actually on 2. Voltage divider after the EF86 to dump signal might be the next step?. There is a lot of bass in the tone so I might try a smaller coupling cap between the first gain stage and tonestack, there is a 0.2 cap at the moment.

[tubeswell]

There is a lot of bass in the tone so I might try a smaller coupling cap between the first gain stage and tonestack, there is a 0.2 cap at the moment.

before you change the coupling cap, try reducing the Ck on that parallel input stage to 2.2uF or thereabouts (1uF - 4.7uF), and see where that leaves the bass rolloff.

[Alexander]

Hey gents--

lots of experimenting with all manner of small pentodes has shown that an EF86 (and 6AU6, 6SJ7GT too) all put out very wide-band signals with small bypass and coupling caps. For coupling caps, a .0047uF usually does what a .022 does on a 12AX7 type and a Ck cap of 2.2uF to 4.7uf same as a 22uF on a 12AX7... i.e., you get 15Hz LF freq response! Which can really swamp a PI or other stage. Why Matchless etc. uses a Ck of 25uF on an EF86 is beyond me-- I've done simple 5E3 vol/tone stacks after a 1st stage EF86 and used a rotary 'bass control', switching in different Ck values, from unbypassed to 1,2.2,3.3, 4.7uF and by the time you hit 4.7 the bass is VERY full!

Anyway, try out some smaller value bits in there, you may be quite surprised.

Cheers,