The Rude Tube

[bernbrue]

Thanks a lot for your calculation,Guavatone, 
just to clarify, the 6AS6 is the second tube stage, not the input tube (EF86 -> 6AS6 -> E88CC) . The 1K5 resistor is just there to prevent the meter to peg to the right, since it is a 1mA meter only. The 10K pot is supposed to be a variable bias resistor. I´m not shure, but your values look like the 6AS6 to work as a linear class A amplifier. Infact, in this context it is meant to work as a distortion tube. I´ll give it a try anyway. The valve wizard site is fantastic and the pentode curves shown there were actually those I had in mind, especially those where the screen voltage (G2) is lowered to 70V or even 30V ('squashed down effect`).
regards
Bernd

 

[flaheu]

Hey Bernbrue,

The amount of work you did is amazing 
I'm following the thread very closely, and of course I can not myself be a big help, because I'm far from a guru  :-[
Any chance to hear some samples ?
I'm sure it should awake people overhere  People believe with their ears !

 

[bernbrue]

1st updated (link to audio sample)
regards
Bernd

 

[ytsestef]

This should be fun. Maybe if it gets documented i'll build one mono REDD47 (which has always been in my plans) and one stereo RudeTube.
Thanks berbrue!

 

[guavatone]

Have you tried this in place of the EF86?  The cool thing about merlin at valve wizard (get his book) is that he talks about gains stages for more musical purposes.  So you may even want to take the load line below the knee for more distortion. 

Actually, try the resistors I suggested.  But change the screen g2 resistor to a pot.  This will make the curves variable.
... maybe finally ending up with 6 values u like, then using a 6 position switch... or maybe 12 values and and 12P switch.
a 100K-250K linear pot should do the trick.

 

[guavatone]

I am sure you know this., but by adding this tube stage you are flipping the polarity of the signal. 

 

[bernbrue]

Hi,
yes, I realized this when I measured the unit with RMAA. This afternoon I tested various G2 resistance values with a pot. Well, this intensified compression effects, which I already had with my values. I´ll give your values a try and see how the unit behaves. I´ve the frontpanel already made (see first post), so another switch won´t be possible. Thanks a lot for your input.
regards
Bernd

 

[bernbrue]

Hi,
I tested the values Guavatone suggested. 27K for anode and 100K for grid 2. Well, soundwise there is a slight difference but these settings do not produce "bias clipping". Would anyone have a closer look at the tutorial videos in order to recapitulate what I mean.

http://www.youtube.com/watch?v=FjzMEH6-SFs

There are two things I do not really understand. How can a cathode be made negative (-2V), is the cathode resistor achieving this? I´m not shure about the 1K resistor in series with grid 1. I saw this in another tube circuit but do not really understand what it does. Sorry for the noob questions. I´m just on a beginners level, but willing to learn.
regards
Bernd

 

[guavatone]

the 1K is just a grid stopper.  Doing nothing really in most cases.

Ok u are trying to get a Vulture!  I want one too.
It sounds like they are getting odd harmonic dist even harmonic dist or both.

I highly recommend merlin's book since he talks of pentode's odd/even harmonics (designing tube preamps for guitar and bass)  He wrights in a nice combo of practical/musical terms and technical as well.

He dissects the EF 86 setup.

()= in parenthesis  are estimates for the 6AS6 tube values - which may be too high,
you also need to be sure it's biased "correctly" and the screen resistor is correct,
which is why I suggested a temporary pot in place of 139K.
both odd and even harmonic dist,:  Ra=100K (33-47K) biased at -2.5K
For symmetrical  Odd Harmonics: he says Ra=47K (maybe 27K)
For symmetrical even: he says Ra=22K (10K or less)

Then try a 500K pot form the the .22uF cap to ground for compression

 

[bernbrue]

Ok u are trying to get a Vulture!  I want one too.

I thought this was quite obvious. It starts to clip/distort only by applying bias and the overdrive switch is not even engaged. "The Rude Tube" doesn´t do that. Though, when overdrive engaged the bias changes the colour of distortion. Is there another way of achieving distortion besides biasing and overdriving the 6AS6? Maybe a cap in parallel to grid 1 and ground? My unit easily distorts when playing powerchords for example, single notes from a bass for example do not distort the way it is shown in the tutorial video. I´m quite satisfied with the results so far, but the last kick hasn´t been achieved yet.

Is the decoupling cap between 2nd and third valve stage really needed?
regards
Bernd

 

[bernbrue]

Hi,
I changed the load line to 100K for anode and 330K for grid 2. With this loadline there is exactly 1mA going through the cathode so that the 1K5 resitor taming the meter became obsolete. Now only the 10K pot is changing the bias. Fine so far. At the weekend I tried various things to increase distortion and emphasize the peak modes. When applying more current to the cathode the peak filters really start to resonant. Would it be allright to tie grid 2 directly to cathode instead of to ground via coupling cap ? (See schematic attached).
regards
Bernd

 

[bernbrue]

That was a bad idea. This loadline also reduces the peak mode effect in such a way that it is almost inaudible. I´m stuck .....

 

[guavatone]

Yeah.  That high loadline will be bad.  I am very curious if the Anode resistor being very low like between 1k-5k will help you get some nice distortion.

By the way, what are you trying to accomplish at this point?  More symmetrical clipping?

 

[bernbrue]

I would like to have a more resonant peak effect and a more "biting" distortion, whatever that means. The variable bias is supposed to produce clipping, but the effect in my unit is too subtle for me. It could do better. G3 concerns me a little bit. I fed signal from various points of the preamp into G3. Well, distortion changes a little bit, but it doesn´t convince me, especially because it isn´t not dynamic, in other words it´s a permanent distortion that is not controlable via gain pot. Yes, Guavatone, you´re right. I had a loadline with 1K5 and 5K6 resistors which worked much better. I´m going to restore it. Are there any other "tricks" to achieve an improvement? Any contribution is very much appreciated.
regards
Bernd

 

[guavatone]

It would be great to hear a sample.  I am very curious of changing the EF86 so that the gain is higher and slams the 6As6
harder to overdrive it more.  Something like the Vox's early use of the EF86 which Merlin warns against in his book.

So try this.

before, I hope you are using a winged C or cheap EF86 you don't mind loosing.....

1. Bypass R7 the 50K res to feed the EF86 with 290V, and listen, adjusting drive
.... by the way, I am concerned about the drive section.  I would just use a 1M pot... at least to see what a higher gain EF86 does.

2.  Play around with R2=220K, R3=1M, R4=2.2K

3. If it's too much replace r7 to  be 22K

 

[bernbrue]

Hi Guavatone,
10K and 39K are great load line settings for the 6AS6. I tried them yesterday. I had to tame the meter with a 2K5 resitor and this actually let the peak modes start to resonant in a pleasant way. I´ll test a resistor between inductor and ground.

Distortion: I think the values for EF86 are quite "conservative". I´d like to push it a little bit further. Using an 1M pot for the drive input section might be a good idea.
I´ll be back soon.
regards
Bernd

 

[bernbrue]

It would be great to hear a sample.  I am very curious of changing the EF86 so that the gain is higher and slams the 6As6
harder to overdrive it more.  Something like the Vox's early use of the EF86 which Merlin warns against in his book.

So try this.

before, I hope you are using a winged C or cheap EF86 you don't mind loosing.....

1. Bypass R7 the 50K res to feed the EF86 with 290V, and listen, adjusting drive
.... by the way, I am concerned about the drive section.  I would just use a 1M pot... at least to see what a higher gain EF86 does.

2.  Play around with R2=220K, R3=1M, R4=2.2K

3. If it's too much replace r7 to  be 22K

Ok, this would also change H.T. for the 6AS6 and its loadline. I´ll try a 1Meg pot for the drive section first. I had a look at the EF86 datasheet and found various values for Ra and Rg. With the REDD stock setting amplification factor of the EF86 is quite low (for less microphonic, I suppose).
regards
Bernd

 

[bernbrue]

OK, now it sounds as I like it.

I lowered R7 to 20K
changed R4 to 2K8
changed drive pot to 1M and pad in front of the pot to 3x 330K
changed 6AS6 loadline to 100K and 330K
rewired Edcor XS1100 to 4:1

That´s the final version 
schematic updated (1st post)
regards
Bernd

 

[guavatone]

Cool.  Is the sample on page one new?  You must be sick of messing around.  I would be

Or, Does it sound good?  Live the CV?

Are you using half of the E88cc  for each channel like CV?

 

[bernbrue]

Yes, I´m really sick of messing around. You should see the board underneath, a mess! I´ll do the second channel and record some audio samples. Well, it sounds "ruuuuuuuuuuude". ;D  Each channel has a dedicated output tube (both triodes in parallel). If anyone ever wants to do a pcb for this, I´m the first who wants one.

Here a kind of step by step guide:
- built a REDD/Orange preamp (stock, except the changed resistor values R4 and R7)
- leave out the NFB loop (leave out the gain switch)
- lift one side of R8 (330K)
- built the distortion valve stage on veroboard and insert it between EF86 and ECC88.
- a few pots and meters here and there, that´s it

regards
Bernd