The Rude Tube

[bernbrue]

Hi Kingston,
I intend to use the Rude Tube exclusively as line level device. Since I already have a Redd47, I won´t need another one. Does the EF86 have to see a step up transformer (1:7 in original circuit) or would a 1:1 transformer with a corresponding pad network make the tube happy as well? A 1:7 transformer would add another 20 dB that I really do not need. For test purpose I took an Edcor 10K/10K, certainly not the best choice, but there are several Lundahls and Sowters in my box waiting to do their job.
regards
Bernd

 

[Kingston]

10K/10K input is quite optimal. A high impedance line input when used with EF86 grids. Does nothing but balanced to unbalanced conversion and galvanic isolation, and the edcor is fine there. I've seen similar solutions recommended for LA2A for taming the very high input levels. Equivalent lundahls and sowters are very expensive.

 

[bernbrue]

10K/10K input is quite optimal. A high impedance line input when used with EF86 grids. Does nothing but balanced to unbalanced conversion and galvanic isolation, and the edcor is fine there. I've seen similar solutions recommended for LA2A for taming the very high input levels. Equivalent lundahls and sowters are very expensive.

Hi Kingston,
thanks for verifying my plan. I have these Lundahls and Sowters in my transformer box. LL1521 (almost similar to the LL1540) will be my first choice. I´ll be back soon.
regards
Bernd

 

[bernbrue]

Hi,
http://twin-x.com/groupdiy/displayimage.php?album=lastup&cat=0&pos=0

for download:

http://twin-x.com/groupdiy/albums/userpics/cv_normal_drive_overdrive.mp3

you are listening to a bad guitar player with a cheap guitar, recorded via DI, processed with "The Rude Tube". First you hear the untreated track, then the RT "normal setting", RT "drive setting" and finally RT "overdrive setting". There are anoying plops when switching through the different modes (Triode, Pentode, Peak 1-4). Is there a solution for this? Enjoy it!
regards
Bernd

 

[mich]

Very nice work -
the dist. sounds good, but it would be nice to hear a drum track or piano or vox with a more subtle distortion setting.
as  guitar distortion tones are easy to get using just some fx pedals...
regarding the pops -
if your switches are 'break before make" - wich i guess they are(?) - you could try a "make before break" switch
so the second grid doesnt hang in the air while switching.
hope it makes sense..

 

[bernbrue]

Hi,
the soundclip was only meant to be a demonstration, that "The Rude Tube" is extremely close to this:

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

Thanks for the advice concerning the switches. I use ordinary 2x6 Lorlins. Is there a "make before brake" (shorting) version as well? Never thought about that.
regards
Bernd

 

[bernbrue]

Hi,
I thought about the peak modes a while. Wouldn´t it be better in pentode mode to connect G2 through the decoupling cap (0,22uF) to cathode instead of to ground? I can imagine that the peak modes would profit by this. What do you think?
regards
Bernd

 

[bernbrue]

ok, I continue my monologue .......
grid 2 to ground or to cathode makes no real difference. The overall frequency response is somewhat lacking in highs. I´m not shure about the 0,47uF cap in front of the 6AS6 tube. Could there be another reason for the strange frequency response?
regards
Bernd

 

[Kingston]

The overall frequency response is somewhat lacking in highs. I´m not shure about the 0,47uF cap in front of the 6AS6 tube. Could there be another reason for the strange frequency response?

The 0.47uF cap between V1 plate and V2 grid is only for DC removal. It only affects bass response, does nothing for high frequencies.

What do you mean lacking in highs? how much? is there a difference between balanced vs. unbalanced out?

And yes, make before break should fix those pops. In a project like this I would hesitate to use those plastic lorlins in the first place. Maybe try to find a better quality Elma or Grayhill 2x6 MBB rotary?

 

[ruffrecords]

There are anoying plops when switching through the different modes (Triode, Pentode, Peak 1-4). Is there a solution for this? Enjoy it!
regards
Bernd

For s1b try 4M7 resistor across each pair of contacts - this ensure all the caps are charged up to the same dc level.

Cheers

Ian

 

[bernbrue]

Hi,
thanks, Ian, for the advice with the 4M7 resistor. I´ll try the other Lorlins first and will solder in the resistors, if the plops are still there. Well, the high frequency roll off is quite strong, -7,5 dB starting from about 5KHz up to 10KHz. I tested it without I/O transformers. AFAIK I didn´t have this roll off before I removed the overall feedback loop. I have to check the EF86 cathode bypass section again, maybe there is a short when I solderd in the 500R trimmer. Here´s the original schematic for reference. 

http://www.sowter.co.uk/schematics/redd47.gif

I have to get rid of all those flying leads and crocodile clips ...
kind regards
Bernd

 

[Kingston]

Well, the high frequency roll off is quite strong, -7,5 dB starting from about 5KHz up to 10KHz. I tested it without I/O transformers.

How long is your unbalanced output testing cable? It might be that the circuit is not at all happy driving even this simple cable, and the cable capacitance is eating your high frequencies, just like happens often with guitars, stomps and amps.

Notice the very high 7:1 ratio of the original output transformer. That's the line driver basically. With something similar you could at least eliminate cable from the equation.

Also, in case you are worried you connected something wrong with the EF86, you could simplify it further and connect it as a triode. See NYD "two bottle" schematic for the (simple) wiring.

By the way, the 220nF cap right after the DI input is unnecessary since the input grid is not in some elevated potential. Unless maybe you are somehow worried you will input DC through the DI. But that should never happen.

 

[kepeb]

any notable improvements since your final sir?

how is it working for you on vox /drums etc?

 

[Kingston]

bernbrue, are you still struggling with the treble roll-off?

See here, your predicament almost exactly:

http://www.groupdiy.com/index.php?topic=44943.msg570724#msg570724

A good cure is setting that first pentode stage into a triode. You lose marginal gain, but you already have too much in the first place. Do it like NYD "two-bottle" front end. Perfectly compatible with EF86. 5879 is interchangeable with it, see below.

 

[bernbrue]

Hi Kingston,
yes, that´s what I thought. Wiring the EF86 as triode could be a solution, but this would mean cutting traces, since my REDD is built on a pcb. What would happen, if I rebuilt the NFB loop? Would distortion, produced by the 6AS6 be obsolete? Is there another way of compensating the treble roll-off? Maybe by changing values in the cathode bypass section? Time is limited at the moment. Anyway, stay tuned.
regards
Bernd

 

[bernbrue]

any notable improvements since your final sir?

how is it working for you on vox /drums etc?

A short summary:
the unit works as shown in the schematic. Results are more than convincing.

Things to to:
- compensating the treble roll-off caused by removing the overall NFB.
- finding suitable I/O transformers
- recalculating input pad for line level

That´s it. l only have built one channel so far. Soundclips will come, when the second channel is built, everything is checked twice and confirmed.
regards
Bernd

 

[bernbrue]

Hi,
wiring the EF86 as triode can be done without cutting traces. Just remove R3, solder a wire from G2 to EF86´s anode (at the bottom of R2). I´m going to try that.
regards
Bernd

 

[bernbrue]

bad news first:
- the treble roll-off is still there 
- the EF86 in triode mode can handle line level easily without pad network 

I´ve just phoned with my mate Reiner who suggested to test signal after each tube stage in order to find the reason for the treble roll off. I´ll do that soon.
regards
Bernd

 

[kepeb]

yea depending on where the treble drops out...
could it be the 3 parallel resistors at the input reducing the impedance?

 

[Kingston]

wiring the EF86 as triode can be done without cutting traces. Just remove R3, solder a wire from G2 to EF86´s anode (at the bottom of R2).

Also need remove C3 and that cathode-to-grid wire. Did you do that? It should look exactly like that two bottle front end pic above. Same pins too if I remember correctly.