Glowing EL84 in Two Rock (and hello)

[HiGain]

Hello Everyone. I've been lurking for a while (what a great place! ), and decided to jump in the game.

I bought a Two Rock 12/18 Ruby a few weeks ago. I made a few mods, listed below:

1.Converted the bright switch to a 100/500pF Cent Off.
2.Added a treble cap switch (47pF/250pF/500pF)
3.Added a treble pot/slope R switch. In position 1 you get 100K slope R and the 1M treble pot with a 330K R across it to make the treble pot behave like a 250K pot. In position 2 you get stock (100K R/1M pot). In position three you get 33K slope R and 250K pot.

With these changes you can access many of the TW schematic variations.

4. I also made the 150K R a Center Detent 250K pot, with a 27K resistor going to ground. Center of the pot equals 150K.

5. Added a cross line cut control with 1M twist SPDT sw. and .0033 polystyrene cap.

6. Added the "body" control listed in Kevin O's second book and shown in the Trainflex schematic. I used a 1M pot, with the cold lug going through a 1.0uF 250Volt cap and a 220K R.

(I also redid the wiring, as per Mark's layout, and swapped out some coupling caps)

OK.... First I plan to make a recording of the different variations, which would be a cool way to compare the differences.

BUT... first, getting to my question... When I employ mod #6, the EL84 that is being turned down begins to glow red hot and my signal gates and then goes. The bias is at minimum setting and reads -18.7. The current at the 18W setting is 52mA, and at the 12W setting is 23mA.

Could someone out there please help me understand why the tube that is receiving less signal overheats?

Thank you!

Jake

[Mark]

Dear Hi Gain

I did have a brief look at the Trainflex circuit and there is a problem. This is basically it. The body pot removes the bias and the valve then runs flat out thus glowing red. Basically you need another decoupling cap so you don't lose your negative bias voltage.

[HiGain]

Hi Mark,

You are correct! I had the 1M pot between the 220K resistor and the tube. I will move the 220K R to bypass the 1M pot.

Thanks for the help!

Jake

[Mark]

Dear Hi Gain

The problem was when the body pot was turned to zero the wiper on the grid was attached to earth thus any negative bias voltage on the grid was eliminated and the tube is running red hot.

The earth side of the body pot should be attach to the bias voltage which is virtually an earth point. Sorry about the vague explanation, I'm doing what I can in very limited time.

[HiGain]

Hi Mark,

No worries... my post in retrospect was a little long for a simple question (just a little excited). However, you pointed me in the right direction. Thanks again...

In the next few weeks I'll try to record some sound samples.

Jake

[Craig Reid]

Hi Gain

You have done something I have toyed with for a while, build a Wreck that covers all of the schematic variations.

Can you tell us how you implemented the switches? Ie., how you wired the treble pot/slope resistor switches and what type of switch you used for each?

I am plotting building a wreck variant with all of the above plus Power Scaling and the Ultimate MV from TUT3. That could potentially be a cool amp.

Craig

[HiGain]

Hi Craig,

No problem...

I have three panel-mounted switches, arranged vertically, and toggling from right to left. I made it so that the center position of all switches is stock as per the A1a schematic.

Bright sw. is just the KF SPDT Cent OFF with 100 and 500pF.

Treble cap sw: I have a 500pF cap on the board, as per the trainwreck layouts. I then have a SP3T switch mounted on the panel. The 500pF mica cap wire goes to the "in" lug of the switch. The lug corresponding to the center position goes directly to the treble pot (meaning the capacitance in this setting is still 500pF). The lug corresponding to the right position goes through another 500pF mica cap (creating a capacitance of 250pF) to the treble pot. The lug corresponding to the left position goes through a 51pF polystyrene cap (equaling around 46pF) to the treble pot. So from left to right, the switch goes 46pF, 500pF, 250pF.

Treble pot/ slope sw: Use a DPDT Cent Off Sw. Center position is stock, meaning nothing is altered. The left position switches a 330K resistor across the outer lugs of the treble pot, making it behave as a 250K pot. The right position does the same thing, while also using the other pole to switch in a parallel 56K slope resistor (forming a slope R of about 36K).

With these switches you get (treble cap/treble pot/slope R) 46pF/250K/100K; 46pF/1M/100K; 46pF/ 250K/36K; 500pF/250K/100K; 500pF/1M/100K; 500pF/250K/36K; 250pF/250K/100K; 250pF/1M/100K; 250pF/250K/36K

You could add 5pF to the 51pF cap on the SP3T sw. to get closer to a net 50pF in the left position, but I actually like the way it sounds as is.

Leaving the bright switch alone, you get a wide range of tones with just these two switches. Right now I am digging 46pF/250K/100K, with a .002/~80K combo.

Hope this was clear!

Jake

[Mark]

Dear Hi Gain

Was your problems with the tube glowing red resolved?

[HiGain]

Hi Mark,

Yes, the problem was resolved. You were correct. I had the 1M pot between the 220K bias resistor and the tube. I rewired the 220K resistor and the problem is gone. Thanks again!

One thing I noticed is that the sweetness and fullness I got when I adjusted the pot before is now gone. I suppose having that extra current run through the tube made it sound extra good! Too bad that would kill the tube.

Jake

[Vince]

Dear Hi Gain

I did have a brief look at the Trainflex circuit ...

Hello for all - my first post here
And - what is "trainflex"?
I check in files - no such schematic

Vince

[Elcabong]

Hello for all - my first post here
And - what is "trainflex"?
I check in files - no such schematic

Vince

It is discussed in that thread

http://maximoaudio.com/forum/viewtopic. ... =trainflex