The Rude Tube

[bernbrue]

Pcbs for the Rude Tube IVORY are available here:

http://groupdiy.com/index.php?topic=53011.0


Hi there,
the Performer has been put on the shelf for a while. The new "The Rude Tube ivory" is on the bench right now. Why always making nice stomp boxes for guitarists? Since I´m a keyboarder and my Motif ES8 has so much space to put things on it I thought of a tube booster for keyboards. It´s a very simple, fender blackface style tube booster with a single russian 6N2P EV for each channel. It´s going to be made p2p and powered by an external stock chinese mic psu. More to come after my holidays.

Update: 1st channel works like charm. Sounds absolutely amazing!!

Update: 2nd channel works as well. I was concerned about the chinese tube mic psu not being beefy enough to provide 2 tubes with current, but it works fine. The CRAZY CHAIN mode is just fantastic. It connects the output of the first channel to the input of the second channel. It distorts like hell. The varible B+ doesn´t work as I thought. Voltage from variable pattern switch is not beefy enough to provide the B+ for two tubes (goes down to 12V or so). The 3 band tone stack is very nice. I´ll keep you up to date.

Update: made some demo tracks

unprocessed file: http://twin-x.com/groupdiy/displayimage.php?pos=-5097

Rude Tube Ivory treated with various Setting in real time while recording: http://twin-x.com/groupdiy/displayimage.php?pos=-5096

For download:

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

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

________________________________________________________________________________________________________
Hi there,
coming sooner or later: The Rude Tube Performer

regards
Bernd

===============================================================================================

Hi,
my REDD to CV mod. Really amazing!

finished unit:

http://www.twin-x.com/groupdiy/albums/userpics/DSCN0013.jpg
http://www.twin-x.com/groupdiy/albums/userpics/DSCN0014.jpg
http://www.twin-x.com/groupdiy/albums/userpics/DSCN0012.jpg

schematic:
http://www.twin-x.com/groupdiy/albums/userpics/CVfinal_tuning.jpg

schematic redrawn for designing the pcb:
http://www.twin-x.com/groupdiy/albums/userpics/RT_-_MB_-_Schematic.pdf

prototype (the beginning):
http://twin-x.com/groupdiy/albums/userpics/TheRudeTube004.jpg

frontpanel:
http://twin-x.com/groupdiy/albums/userpics/TheRudeTube002.jpg

frontpanel fpd file zipped:
http://twin-x.com/groupdiy/albums/userpics/Rude_Tube.zip

The meter cutouts are not very precice, check before having made a frontpanel

audio sample (with switch plops)
http://twin-x.com/groupdiy/albums/userpics/cv_normal_drive_overdrive.mp3

audio sample final version
http://www.twin-x.com/groupdiy/albums/userpics/rude_tube_demo_2.mp3

audio demo with drumloop demonstrating the peak modes and ...... distortion
http://www.twin-x.com/groupdiy/albums/userpics/rude_tube_drums.mp3

regards
Bernd

 

[flaheu]

Amazing !!

Any chance to hear samples  ? 

 

[Kingston]

Great and simple implementation with that pentode and peaking filter switch. Thanks for posting.

But there surely must be a typo with that ECC83 substitution bit! It's 12AX7, and so is 6n2p! I'm surprised it works at all with a big output transformer. Don't you have a massive bass cut quite audible?

The real Russian E88CC equivalent is 6N23P, which is even perfectly pin compatible, and will happily drive lines with something like 2:1 or 4:1 output transformer in that parallel configuration.

Again, 6n2p is basically a 12AX7 and is not even pin compatible! Do not trust those eBay "equivalency" claims. Check datasheets first.

DANGER! DANGER!

In other news, note that 6N1P is often sold as ecc88 equivalent too. It is not. It has higher heater current needs and a weaker plate (will not drive lines as easily). There is no clear 6N1P western equivalent, but it would nevertheless work ok in this configuration with something like a 4:1 output transformer.

 

[bernbrue]

Hi Kingston,
I´ve just started to read about tube theory. The mod was actually done only by trying, just having a vague notion about the theory and the math behind it.

My aim with the substitution of the ECC88 with the russian ECC83 was just to be closer to the real thing but the difference between ECC88 and the russian tube is not really audible.

At the moment I run the prototype without any input/output tx. Concerning impedance and ratio of the input/output tx I´m a little bit lost. Here I need some help. 

The L pad in front of the 100K drive pot tames the huge amount of gain coming from the EF86. Since the level change when engaging the "overdrive" switch is quite big I would like to have three steps of gain reduction. Is there an easy way to implement this?

The peak modes are very interesting and the effect is really nice. It could have a bit more intensity, though. Maybe I should try another inductor value . Suggestions are very much appreciated.


Soundwise the "The Rude Tube" is extremly versatile. From creamy triode sound to heavy, rectifier like pentode distortion.

Build it and you know it.

regards
Bernd

 

[Kingston]

My aim with the substitution of the ECC88 with the russian ECC83 was just to be closer to the real thing but the difference between ECC88 and the russian tube is not really audible.

What real thing? There isn't a single design out there where 12AX7/ECC83 is successfully used as a line driver like ECC88/6DJ8 is used here. It all boils down to designing Common-Cathode amplifier stages, and choosing the right - or at least closely related - tube for the task. I can't explain it better than this or in any short way: http://www.aikenamps.com/CommonCathode.htm

The design parameters:  This last stage should not greatly influence the sound of the already distorted signal. Its only task is to mildly and linearly amplify the signal, while trying to pass enough current for the transformer to drive roughly 600-ohm lines.

please proceed very very carefully here. 6n2p is not even the same pinout as either 12AX7 or ECC88, so it isn't a drop in replacement. There's a good chance your heaters are currently wired wrong and the tubes are taking their last breath. They're cheap, but the error could just as well be something more lethal with these voltages.

In fact, forget about this substitution thing for now and use ECC88 or a direct equivalent. For the more discerning, they are also guaranteed to sound better than a 12AX7 in this driver position.

At the moment I run the prototype without any input/output tx. Concerning impedance and ratio of the input/output tx I´m a little bit lost. Here I need some help.

This explains why it still works, but see the heater comment above why it might not work for too long anymore.

I thought you used the output transformer like in the schematic. That will result in strangeness like bass cuts etc. attenuation. A tube like 12AX7 set up in this very ill way will still (barely) work with modern day high impedance AD converter input, which you are probably using for testing. But this is a happy day, actually lucky day scenario. Connect it to just about anything else (like an output transformer or input of most gear) and you will find out why 12AX7 is never used this way.

The L pad in front of the 100K drive pot tames the huge amount of gain coming from the EF86. Since the level change when engaging the "overdrive" switch is quite big I would like to have three steps of gain reduction. Is there an easy way to implement this?

Yes there is.

The 750+750 ohm L-pad is a not a good idea in the first place. I can't imagine the EF86 being too happy driving this less than 1k load, and it's always on. In fact I'm surprised you don't have strange side effects from having done that, too. This L-pad (in other words a voltage divider) should be in the range of something like 50k-200k for the tube to be relatively happy. In fact you could just use another 100k divider, right next to that "drive" pot.

Here, calculate several steps for a 100k switch, like "no attenuation", -10dB, -20dB, etc.

http://homepages.tcp.co.uk/~nroberts/atten.html

 

[ChrioN]

coolage. keep us updated!

 

[bernbrue]

Thanks a lot for your explanation, Kingston,
I tried an Edcor XS1100 in 2:1 configuration as output tx and the results are fine so far, exact measurements will follow. I also replaced the russian tube with a NOS Valvo E88CC. E-Guitar goes straight into the DI input. I still have to change the 750/750 ohm voltage divider and find a suitable input tx. The peak modes could be more pronounced. Is the 100 uH inductor correct or would another value or configuration be better? In my prototype the output volume is regulated by a 100K pot after the E88CC. I know that it would be better to place is before the output valve, but this would mean cutting traces, which I would like to avoid. Any suggestions are very much appreciated.
regards
Bernd

 

[bernbrue]

Hi,
I´ve updated the schematic in the first post. The overdrive switch has three steps now. As you can see, I really had to bring down a huge amount of gain coming from the EF86. In normal position, drive and output pot in 12 o´clock position, I have unity gain.  With drive pot turned full CW the 6as6 gently starts to distort, even in triode mode. Bias changes the colour of distortion. Overdrive switch in mid position gives you a quite rich distortion depending on drive pot position and chosen mode. Overdrive switch down, pentode or one of the peak modes, drive pot fully cranked up and you´re blown away ....................   Lead guitar sounds fantastic, sustain more than 15 seconds. The peak modes are working fine now (there was a grounding problem).  Sound clips not before Sunday.
regards
Bernd

 

[Kingston]

In my prototype the output volume is regulated by a 100K pot after the E88CC. I know that it would be better to place is before the output valve, but this would mean cutting traces, which I would like to avoid. Any suggestions are very much appreciated.

Having the output volume before the output stage is not automatically better. Now you have a chance to slightly overdrive that stage too, might as well leave it there since it's an overdrive unit at heart anyway. My preferred place in a plan like this would be a balanced 600-ohm T-attenuator after the output transformer. But then you have to make sure the trafo is beefy enough because clipping a transformer usually sounds plain bad, even in a unit like this.

 

[bernbrue]

Hi Kingston,
thanks a lot for your input. I learned a lot from your "Drive1" project and was really motivated to try something similar, though your project is a totally different beast.  All the other "distortion similar to .... " threads were quite disappointing to read (people´s attitude and so on ...).  The more I was encouraged to do it myself.
kind regards
Bernd

 

[Kingston]

These are the type of projects I like best. Glad to be of any use for you. There's just not enough controllable distortion out there.

Couple of additional points and questions.

What is the function of R8? Something to do with taming the pentode output? It seems awfully large, actually unnecessary, but I may have missed something.

R9 is possibly not needed. A parallel ECC88 stage like this is quite linear even when running with no feedback at all. 1M6 ohms of feedback is so very little - probably less than 1dB - that you could just as well remove it.

 

[volker]

No, R8 and R9 work together to give negative feedback. So that sets it to roughly 13dB,
instead of open loop with probably something like 25-30dB.

Additionally to better linearity, it reduces the output impedance of the ECC88 stage.


Great project Bernd!

 

[bernbrue]

What is the function of R8? Something to do with taming the pentode output? It seems awfully large, actually unnecessary, but I may have missed something.

R9 is possibly not needed. A parallel ECC88 stage like this is quite linear even when running with no feedback at all. 1M6 ohms of feedback is so very little - probably less than 1dB - that you could just as well remove it.

When this project started a few month ago, a friend of mine explained to me that the signal from the E88CCs anode is inverted to the grid. R9 sends this inverted signal back and then it is mixed with the signal comming from R8. A kind of inverting opamp. I think I leave it this way. I tried the circuit without R8, but the pentode´s output was almost overdriving the E88CC.  Simply too much gain for a line level distortion unit. By the way, a very nice setting is the triode mode with about 0,2 mA going through the cathode. My ears are pampered with very rich 2nd harmonics. I love it.
regards
Bernd 

 

[tv]

perhaps you could try 100K for R8 and cca. 470K for R9...

 

[bernbrue]

Hi,
tested some input transformers today. The EF86 starts to distort in an ugly way when fed with line level. I tried various 10K/10K input tx, which resulted in unpleasant distortion. If I took input transformers with higher ratio (1:4) there would be only more input gain, So do I need a kind of pad before the EF86, or is the problem caused by an impedance mismatch? The DI input with guitar plugged works fine. I have to look at Jacobs G9 circuit for a solution.
regards
Bernd

 

[bernbrue]

Aha !!!

 

[kepeb]

go on...

 

[kepeb]

the g9 input is 1:6.45 I would think this will still distort as you said earlier.
are you feeding the D.I. straight to the ef86?

congrats by the way, pioneering diy groundwork, thanks for sharing 

 

[bernbrue]

Hi,
updated the schematic (1st post)
regards
Bernd

 

[Kingston]

The pad used in G9 might or might not be optimal with your input transformer. I don't know how much you have already "locked in" your front panel design, but the EF86 (or similar) high gain input stage would easily allow you to use this unit as a mic preamp with padded input vs. non-padded switch. Let's say you had a 1:5 - 1:10 or whatever high step up input transformer. Now you have a very flexible preamp. All you need is something like -30dB H- or U-pad before the input transformer and now you have a line level unit. See my Drive-1 design where I did exactly this with a switch.

to calculate your own optimal input pad:

http://www.nu9n.com/tpad-calculator.html

or here for a whole walkthrough:

http://www.uneeda-audio.com/pads/